| data |
Words |
1493 |
45 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 157 uses |
| ai |
Words |
1214 |
45 |
Attestation across the AI Supply Chain 111 uses |
| model |
Words |
371 |
42 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 70 uses |
| training |
Words |
304 |
41 |
Attestation across the AI Supply Chain 54 uses |
| like |
Words |
283 |
42 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 32 uses |
| people |
Words |
251 |
40 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 28 uses |
| public |
Words |
234 |
37 |
Public AI, Data Appraisal, and Data Debates 48 uses |
| models |
Words |
229 |
41 |
"Many Models" and "Track Changes" for AI: Some Thoughts on LLM Interfaces 19 uses |
| think |
Words |
226 |
43 |
Attestation across the AI Supply Chain 23 uses |
| use |
Words |
215 |
43 |
Attestation across the AI Supply Chain 24 uses |
| work |
Words |
208 |
41 |
Attestation across the AI Supply Chain 28 uses |
| post |
Words |
199 |
43 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 13 uses |
| systems |
Words |
194 |
38 |
Building a Data Pipeworks for Democratic AI: From Human Knowledge to Records to AI Systems 28 uses |
| information |
Words |
191 |
28 |
Attestation across the AI Supply Chain 32 uses |
| just |
Words |
182 |
41 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 21 uses |
| one |
Words |
178 |
42 |
Attestation across the AI Supply Chain 25 uses |
| make |
Words |
176 |
39 |
Attestation across the AI Supply Chain 25 uses |
| new |
Words |
174 |
37 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 15 uses |
| want |
Words |
174 |
40 |
Attestation across the AI Supply Chain 20 uses |
| see |
Words |
169 |
36 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 20 uses |
| content |
Words |
166 |
28 |
Will the New York Times Data Strike Have a Large Impact on ChatGPT? 25 uses |
| many |
Words |
165 |
40 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 18 uses |
| knowledge |
Words |
161 |
28 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 35 uses |
| evaluation |
Words |
158 |
18 |
Attestation across the AI Supply Chain 69 uses |
| users |
Words |
150 |
26 |
Bing Rewards for the AI Age 23 uses |
| value |
Words |
150 |
34 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 18 uses |
| outputs |
Words |
149 |
27 |
Attestation across the AI Supply Chain 54 uses |
| collective |
Words |
147 |
37 |
Algorithmic Collective Action With Two Collectives [crosspost] 30 uses |
| open |
Words |
141 |
26 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 59 uses |
| leverage |
Words |
137 |
28 |
A Short Guide to Data Strikes and Conscious Data Contribution in the Context of 2026 Frontier AI 20 uses |
| good |
Words |
129 |
35 |
Attestation across the AI Supply Chain 17 uses |
| wikipedia |
Words |
128 |
17 |
Bing Rewards for the AI Age 25 uses |
| research |
Words |
124 |
32 |
Evaluation Data Leverage: Advances like "Deep Research" Highlight a Looming Opportunity for Bargaining Power 18 uses |
| get |
Words |
120 |
34 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 15 uses |
| set |
Words |
117 |
33 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 13 uses |
| system |
Words |
116 |
29 |
Bing Rewards for the AI Age 22 uses |
| human |
Words |
115 |
31 |
Building a Data Pipeworks for Democratic AI: From Human Knowledge to Records to AI Systems 21 uses |
| commons |
Words |
114 |
22 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 34 uses |
| likely |
Words |
112 |
36 |
Attestation across the AI Supply Chain 19 uses |
| labor |
Words |
111 |
22 |
Attestation across the AI Supply Chain 20 uses |
| etc |
Words |
109 |
30 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 25 uses |
| now |
Words |
109 |
31 |
The Paradox of Reuse in 2026: A Case of Quasi-Enclosure, or "Subsidized Club Goods that Sort of Look Like Public Goods" 12 uses |
| organizations |
Words |
108 |
28 |
Public AI, Data Appraisal, and Data Debates 15 uses |
| much |
Words |
106 |
35 |
Attestation across the AI Supply Chain 11 uses |
| using |
Words |
105 |
34 |
"Many Models" and "Track Changes" for AI: Some Thoughts on LLM Interfaces 8 uses |
| used |
Words |
104 |
35 |
Attestation across the AI Supply Chain 11 uses |
| time |
Words |
103 |
32 |
Attestation across the AI Supply Chain 9 uses |
| still |
Words |
101 |
30 |
Attestation across the AI Supply Chain 16 uses |
| actually |
Words |
93 |
28 |
The Paradox of Reuse in 2026: A Case of Quasi-Enclosure, or "Subsidized Club Goods that Sort of Look Like Public Goods" 10 uses |
| kind |
Words |
93 |
30 |
Attestation across the AI Supply Chain 13 uses |
| social |
Words |
93 |
27 |
Building a Data Pipeworks for Democratic AI: From Human Knowledge to Records to AI Systems 19 uses |
| chatgpt |
Words |
92 |
23 |
ChatGPT is Awesome and Scary: You Deserve Credit for the Good Parts (and Might Help Fix the Bad Parts) 23 uses |
| need |
Words |
92 |
29 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 11 uses |
| around |
Words |
91 |
34 |
Attestation across the AI Supply Chain 20 uses |
| companies |
Words |
91 |
23 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 32 uses |
| help |
Words |
91 |
28 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 12 uses |
| making |
Words |
91 |
39 |
Attestation across the AI Supply Chain 19 uses |
| useful |
Words |
90 |
28 |
Attestation across the AI Supply Chain 10 uses |
| different |
Words |
89 |
31 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 11 uses |
| going |
Words |
89 |
34 |
Attestation across the AI Supply Chain 8 uses |
| specific |
Words |
89 |
35 |
Building a Data Pipeworks for Democratic AI: From Human Knowledge to Records to AI Systems 8 uses |
| high |
Words |
88 |
29 |
The Paradox of Reuse in 2026: A Case of Quasi-Enclosure, or "Subsidized Club Goods that Sort of Look Like Public Goods" 11 uses |
| quality |
Words |
88 |
23 |
AI is driving the cost of polish down; some musings on fancy versus terse artifacts 18 uses |
| code |
Words |
87 |
21 |
The Coding Agent Data Deal 17 uses |
| impact |
Words |
86 |
28 |
Will the New York Times Data Strike Have a Large Impact on ChatGPT? 14 uses |
| important |
Words |
83 |
32 |
Attestation across the AI Supply Chain 9 uses |
| something |
Words |
83 |
28 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 10 uses |
| create |
Words |
82 |
30 |
Attestation across the AI Supply Chain 11 uses |
| large |
Words |
82 |
34 |
Will the New York Times Data Strike Have a Large Impact on ChatGPT? 6 uses |
| really |
Words |
82 |
31 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 10 uses |
| user |
Words |
81 |
18 |
The Coding Agent Data Deal 23 uses |
| action |
Words |
80 |
22 |
Algorithmic Collective Action With Two Collectives [crosspost] 18 uses |
| certain |
Words |
80 |
26 |
Attestation across the AI Supply Chain 14 uses |
| coding |
Words |
80 |
12 |
The Coding Agent Data Deal 41 uses |
| way |
Words |
80 |
34 |
Attestation across the AI Supply Chain 7 uses |
| capabilities |
Words |
79 |
23 |
AI Dividends Without Taxing Compute, Automation, or Equity: A Presumptive Commons-Rent Tax Based on Capabilities and Data Dependence 11 uses |
| output |
Words |
79 |
18 |
Attestation across the AI Supply Chain 40 uses |
| platforms |
Words |
79 |
19 |
Bing Rewards for the AI Age 16 uses |
| access |
Words |
78 |
24 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 11 uses |
| agents |
Words |
78 |
12 |
The Coding Agent Data Deal 30 uses |
| bargaining |
Words |
78 |
24 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 12 uses |
| rules |
Words |
78 |
13 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 47 uses |
| world |
Words |
78 |
27 |
Building a Data Pipeworks for Democratic AI: From Human Knowledge to Records to AI Systems 13 uses |
| agent |
Words |
76 |
13 |
The Coding Agent Data Deal 39 uses |
| case |
Words |
76 |
36 |
Live by the free-content-for-training sword, die by the free-content-for-training sword 6 uses |
| dataset |
Words |
76 |
25 |
Attestation across the AI Supply Chain 11 uses |
| individual |
Words |
76 |
26 |
Bing Rewards for the AI Age 12 uses |
| openai |
Words |
76 |
26 |
ChatGPT is Awesome and Scary: You Deserve Credit for the Good Parts (and Might Help Fix the Bad Parts) 10 uses |
| related |
Words |
76 |
23 |
Attestation across the AI Supply Chain 12 uses |
| llm |
Words |
75 |
21 |
Bing Rewards for the AI Age 17 uses |
| better |
Words |
73 |
27 |
The Paradox of Reuse in 2026: A Case of Quasi-Enclosure, or "Subsidized Club Goods that Sort of Look Like Public Goods" 10 uses |
| course |
Words |
73 |
29 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 8 uses |
| search |
Words |
73 |
23 |
Bing Rewards for the AI Age 16 uses |
| current |
Words |
72 |
26 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 12 uses |
| lot |
Words |
72 |
27 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 7 uses |
| generative |
Words |
71 |
17 |
Google and TikTok rank bundles of information; ChatGPT ranks grains. 12 uses |
| power |
Words |
71 |
26 |
On AI-driven Job Apocalypses and Collective Bargaining for Information 11 uses |
| terms |
Words |
71 |
24 |
Live by the free-content-for-training sword, die by the free-content-for-training sword 10 uses |
| already |
Words |
70 |
26 |
Attestation across the AI Supply Chain 10 uses |
| probably |
Words |
70 |
25 |
Will the New York Times Data Strike Have a Large Impact on ChatGPT? 8 uses |
| well |
Words |
70 |
34 |
Evaluation Data Leverage: Advances like "Deep Research" Highlight a Looming Opportunity for Bargaining Power 6 uses |
| provenance |
Words |
69 |
8 |
Attestation across the AI Supply Chain 29 uses |
| creators |
Words |
68 |
17 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 28 uses |
| part |
Words |
68 |
26 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 12 uses |
| perhaps |
Words |
68 |
27 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 7 uses |
| first |
Words |
67 |
34 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 8 uses |
| idea |
Words |
67 |
31 |
AI Dividends Without Taxing Compute, Automation, or Equity: A Presumptive Commons-Rent Tax Based on Capabilities and Data Dependence 6 uses |
| level |
Words |
67 |
30 |
Tipping Points for Content Ecosystems 7 uses |
| support |
Words |
67 |
27 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 12 uses |
| another |
Words |
66 |
31 |
Attestation across the AI Supply Chain 8 uses |
| reddit |
Words |
66 |
13 |
Reddit, StackOverflow, and Europe: All Trending Towards Data Dignity 19 uses |
| though |
Words |
66 |
27 |
The WGA Strike is a Canary in the Coal Mine for AI Labor Concerns 7 uses |
| attestation |
Words |
65 |
5 |
Attestation across the AI Supply Chain 57 uses |
| contributions |
Words |
65 |
21 |
Bing Rewards for the AI Age 11 uses |
| point |
Words |
65 |
33 |
Tipping Points for Content Ecosystems 6 uses |
| products |
Words |
65 |
18 |
Attestation across the AI Supply Chain 14 uses |
| right |
Words |
65 |
24 |
The WGA Strike is a Canary in the Coal Mine for AI Labor Concerns 11 uses |
| sharing |
Words |
65 |
23 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 11 uses |
| web |
Words |
65 |
24 |
The Coding Agent Data Deal 11 uses |
| become |
Words |
64 |
30 |
Attestation across the AI Supply Chain 14 uses |
| strike |
Words |
64 |
8 |
Will the New York Times Data Strike Have a Large Impact on ChatGPT? 25 uses |
| tools |
Words |
64 |
23 |
Coding agents are (1) a big deal, (2) very relevant to data leverage, and (3) able to help build tools that support data leverage! 18 uses |
| approach |
Words |
63 |
27 |
Attestation across the AI Supply Chain 8 uses |
| feedback |
Words |
63 |
21 |
Bing Rewards for the AI Age 16 uses |
| labs |
Words |
63 |
15 |
Attestation across the AI Supply Chain 14 uses |
| questions |
Words |
63 |
26 |
Building a Data Pipeworks for Democratic AI: From Human Knowledge to Records to AI Systems 7 uses |
| market |
Words |
62 |
12 |
Attestation across the AI Supply Chain 19 uses |
| markets |
Words |
62 |
12 |
Attestation across the AI Supply Chain 32 uses |
| note |
Words |
62 |
21 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 10 uses |
| points |
Words |
62 |
21 |
Tipping Points for Content Ecosystems 10 uses |
| source |
Words |
62 |
22 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 17 uses |
| based |
Words |
61 |
29 |
Attestation across the AI Supply Chain 10 uses |
| flow |
Words |
61 |
21 |
Plural AI Data Alignment 8 uses |
| paper |
Words |
61 |
22 |
ChatGPT is Awesome and Scary: You Deserve Credit for the Good Parts (and Might Help Fix the Bad Parts) 6 uses |
| policy |
Words |
61 |
21 |
AI Dividends Without Taxing Compute, Automation, or Equity: A Presumptive Commons-Rent Tax Based on Capabilities and Data Dependence 9 uses |
| particular |
Words |
60 |
29 |
A Short Guide to Data Strikes and Conscious Data Contribution in the Context of 2026 Frontier AI 6 uses |
| short |
Words |
60 |
31 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 7 uses |
| workers |
Words |
60 |
11 |
On AI-driven Job Apocalypses and Collective Bargaining for Information 12 uses |
| however |
Words |
59 |
32 |
Reddit, StackOverflow, and Europe: All Trending Towards Data Dignity 4 uses |
| incentives |
Words |
59 |
15 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 15 uses |
| llms |
Words |
59 |
20 |
Bing Rewards for the AI Age 13 uses |
| via |
Words |
59 |
25 |
Attestation across the AI Supply Chain 9 uses |
| future |
Words |
58 |
28 |
The AI "Evaluation Crisis" Is an Opportunity to Get Data Flow Right 6 uses |
| credit |
Words |
57 |
11 |
Bing Rewards for the AI Age 23 uses |
| across |
Words |
56 |
24 |
Attestation across the AI Supply Chain 15 uses |
| general |
Words |
56 |
26 |
Attestation across the AI Supply Chain 13 uses |
| instance |
Words |
56 |
27 |
The Coding Agent Data Deal 5 uses |
| term |
Words |
56 |
26 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 10 uses |
| try |
Words |
56 |
27 |
AI progress as quasi-public good production 6 uses |
| two |
Words |
56 |
26 |
The Paradox of Reuse in 2026: A Case of Quasi-Enclosure, or "Subsidized Club Goods that Sort of Look Like Public Goods" 5 uses |
| back |
Words |
55 |
21 |
The Paradox of Reuse in 2026: A Case of Quasi-Enclosure, or "Subsidized Club Goods that Sort of Look Like Public Goods" 6 uses |
| groups |
Words |
55 |
17 |
Plural AI Data Alignment 10 uses |
| let |
Words |
55 |
24 |
Tipping Points for Content Ecosystems 8 uses |
| number |
Words |
55 |
28 |
AI is driving the cost of polish down; some musings on fancy versus terse artifacts 5 uses |
| records |
Words |
55 |
16 |
Building a Data Pipeworks for Democratic AI: From Human Knowledge to Records to AI Systems 16 uses |
| say |
Words |
55 |
28 |
Attestation across the AI Supply Chain 6 uses |
| contribution |
Words |
54 |
20 |
Tipping Points for Content Ecosystems 10 uses |
| credits |
Words |
54 |
9 |
Bing Rewards for the AI Age 44 uses |
| economic |
Words |
54 |
25 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 7 uses |
| know |
Words |
54 |
29 |
Data Leverage Recap: December 2022 - April 2023 6 uses |
| consider |
Words |
53 |
25 |
A Short Guide to Data Strikes and Conscious Data Contribution in the Context of 2026 Frontier AI 9 uses |
| directly |
Words |
53 |
27 |
Bing Rewards for the AI Age 10 uses |
| end |
Words |
53 |
23 |
The Paradox of Reuse in 2026: A Case of Quasi-Enclosure, or "Subsidized Club Goods that Sort of Look Like Public Goods" 9 uses |
| key |
Words |
53 |
23 |
Data Leverage Recap: December 2022 - April 2023 8 uses |
| produce |
Words |
53 |
27 |
The WGA Strike is a Canary in the Coal Mine for AI Labor Concerns 6 uses |
| towards |
Words |
53 |
25 |
AI Labs Should Open Source Data Protection Technologies 6 uses |
| without |
Words |
53 |
27 |
AI is driving the cost of polish down; some musings on fancy versus terse artifacts 6 uses |
| blog |
Words |
52 |
24 |
ChatGPT is Awesome and Scary: You Deserve Credit for the Good Parts (and Might Help Fix the Bad Parts) 10 uses |
| give |
Words |
52 |
25 |
Bing Rewards for the AI Age 5 uses |
| posts |
Words |
52 |
20 |
How I am publishing right now 9 uses |
| problem |
Words |
52 |
17 |
AI is driving the cost of polish down; some musings on fancy versus terse artifacts 12 uses |
| production |
Words |
52 |
19 |
The Paradox of Reuse in 2026: A Case of Quasi-Enclosure, or "Subsidized Club Goods that Sort of Look Like Public Goods" 13 uses |
| take |
Words |
52 |
27 |
Attestation across the AI Supply Chain 6 uses |
| google |
Words |
51 |
21 |
Google and TikTok rank bundles of information; ChatGPT ranks grains. 7 uses |
| train |
Words |
51 |
28 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 6 uses |
| cases |
Words |
50 |
26 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 7 uses |
| full |
Words |
50 |
25 |
Attestation across the AI Supply Chain 9 uses |
| learning |
Words |
50 |
18 |
The AI "Evaluation Crisis" Is an Opportunity to Get Data Flow Right 13 uses |
| long |
Words |
50 |
26 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 10 uses |
| potential |
Words |
50 |
23 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 5 uses |
| process |
Words |
50 |
17 |
Attestation across the AI Supply Chain 9 uses |
| provide |
Words |
50 |
20 |
Bing Rewards for the AI Age 7 uses |
| share |
Words |
50 |
21 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 11 uses |
| trying |
Words |
50 |
28 |
Which datasets should we assume are "in all the AI models"? 4 uses |
| build |
Words |
49 |
22 |
Building a Data Pipeworks for Democratic AI: From Human Knowledge to Records to AI Systems 5 uses |
| especially |
Words |
49 |
29 |
Building a Data Pipeworks for Democratic AI: From Human Knowledge to Records to AI Systems 5 uses |
| group |
Words |
49 |
21 |
A Harbinger of the Future of Content? The New York Times Starts a Data Strike 7 uses |
| ways |
Words |
49 |
23 |
A Short Guide to Data Strikes and Conscious Data Contribution in the Context of 2026 Frontier AI 6 uses |
| building |
Words |
48 |
18 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 8 uses |
| given |
Words |
48 |
26 |
Plural AI Data Alignment 5 uses |
| governance |
Words |
48 |
14 |
Plural AI Data Alignment 6 uses |
| often |
Words |
48 |
27 |
Attestation across the AI Supply Chain 8 uses |
| various |
Words |
48 |
23 |
On AI-driven Job Apocalypses and Collective Bargaining for Information 5 uses |
| argument |
Words |
47 |
23 |
Building a Data Pipeworks for Democratic AI: From Human Knowledge to Records to AI Systems 6 uses |
| example |
Words |
47 |
19 |
Attestation across the AI Supply Chain 13 uses |
| issues |
Words |
47 |
18 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 14 uses |
| look |
Words |
47 |
20 |
Attestation across the AI Supply Chain 6 uses |
| major |
Words |
47 |
22 |
A Harbinger of the Future of Content? The New York Times Starts a Data Strike 6 uses |
| progress |
Words |
47 |
18 |
AI is driving the cost of polish down; some musings on fancy versus terse artifacts 7 uses |
| technical |
Words |
47 |
22 |
Attestation across the AI Supply Chain 9 uses |
| benefit |
Words |
46 |
22 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 13 uses |
| context |
Words |
46 |
22 |
[microblog] One book is worth "0.06%" benchmark points to AI; is "no different from noise". What gives? 4 uses |
| keep |
Words |
46 |
23 |
On AI-driven Job Apocalypses and Collective Bargaining for Information 5 uses |
| means |
Words |
46 |
24 |
Attestation across the AI Supply Chain 6 uses |
| private |
Words |
46 |
13 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 12 uses |
| rather |
Words |
46 |
19 |
Attestation across the AI Supply Chain 12 uses |
| similar |
Words |
46 |
20 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 8 uses |
| tech |
Words |
46 |
17 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 11 uses |
| transparency |
Words |
46 |
15 |
Attestation across the AI Supply Chain 9 uses |
| writing |
Words |
46 |
21 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 7 uses |
| efforts |
Words |
45 |
21 |
The AI "Evaluation Crisis" Is an Opportunity to Get Data Flow Right 8 uses |
| product |
Words |
45 |
13 |
Attestation across the AI Supply Chain 21 uses |
| run |
Words |
45 |
21 |
The Coding Agent Data Deal 5 uses |
| version |
Words |
45 |
20 |
Attestation across the AI Supply Chain 18 uses |
| design |
Words |
44 |
17 |
Attestation across the AI Supply Chain 10 uses |
| possible |
Words |
44 |
22 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 8 uses |
| text |
Words |
44 |
12 |
The AI "Evaluation Crisis" Is an Opportunity to Get Data Flow Right 11 uses |
| things |
Words |
44 |
22 |
How do we know our AI output is good? Double checks, bar charts, vibes, and training data. 5 uses |
| firms |
Words |
43 |
15 |
Bing Rewards for the AI Age 11 uses |
| ideas |
Words |
43 |
18 |
"People First" Policy Ideas that Complement Each Other (through better data flow) 11 uses |
| made |
Words |
43 |
27 |
Attestation across the AI Supply Chain 6 uses |
| small |
Words |
43 |
21 |
Attestation across the AI Supply Chain 8 uses |
| specifically |
Words |
43 |
26 |
Building a Data Pipeworks for Democratic AI: From Human Knowledge to Records to AI Systems 5 uses |
| artifacts |
Words |
42 |
13 |
AI is driving the cost of polish down; some musings on fancy versus terse artifacts 23 uses |
| communities |
Words |
42 |
13 |
AI is driving the cost of polish down; some musings on fancy versus terse artifacts 11 uses |
| existing |
Words |
42 |
18 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 7 uses |
| quite |
Words |
42 |
26 |
The AI "Evaluation Crisis" Is an Opportunity to Get Data Flow Right 4 uses |
| recent |
Words |
42 |
18 |
Will the New York Times Data Strike Have a Large Impact on ChatGPT? 6 uses |
| seems |
Words |
42 |
21 |
Will the New York Times Data Strike Have a Large Impact on ChatGPT? 5 uses |
| working |
Words |
42 |
24 |
Attestation across the AI Supply Chain 5 uses |
| claude |
Words |
41 |
10 |
The Coding Agent Data Deal 13 uses |
| datasets |
Words |
41 |
19 |
Building a Data Pipeworks for Democratic AI: From Human Knowledge to Records to AI Systems 9 uses |
| experiments |
Words |
41 |
15 |
Public AI, Data Appraisal, and Data Debates 13 uses |
| generally |
Words |
41 |
20 |
Attestation across the AI Supply Chain 9 uses |
| instead |
Words |
41 |
21 |
Attestation across the AI Supply Chain 6 uses |
| involve |
Words |
41 |
21 |
Attestation across the AI Supply Chain 7 uses |
| legal |
Words |
41 |
14 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 9 uses |
| tasks |
Words |
41 |
17 |
On AI-driven Job Apocalypses and Collective Bargaining for Information 9 uses |
| times |
Words |
41 |
17 |
A Harbinger of the Future of Content? The New York Times Starts a Data Strike 11 uses |
| upstream |
Words |
41 |
16 |
Attestation across the AI Supply Chain 16 uses |
| answer |
Words |
40 |
17 |
The Paradox of Reuse, Language Models Edition 7 uses |
| collectives |
Words |
40 |
5 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 16 uses |
| discussions |
Words |
40 |
21 |
Building a Data Pipeworks for Democratic AI: From Human Knowledge to Records to AI Systems 6 uses |
| every |
Words |
40 |
22 |
How do we know our AI output is good? Double checks, bar charts, vibes, and training data. 6 uses |
| hard |
Words |
40 |
23 |
On AI-driven Job Apocalypses and Collective Bargaining for Information 5 uses |
| technologies |
Words |
40 |
19 |
Data Leverage Recap: December 2022 - April 2023 6 uses |
| thinking |
Words |
40 |
19 |
Building a Data Pipeworks for Democratic AI: From Human Knowledge to Records to AI Systems 6 uses |
| big |
Words |
39 |
17 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 14 uses |
| details |
Words |
39 |
19 |
Which datasets should we assume are "in all the AI models"? 4 uses |
| discussion |
Words |
39 |
22 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 8 uses |
| goods |
Words |
39 |
8 |
The Paradox of Reuse in 2026: A Case of Quasi-Enclosure, or "Subsidized Club Goods that Sort of Look Like Public Goods" 14 uses |
| jobs |
Words |
39 |
10 |
On AI-driven Job Apocalypses and Collective Bargaining for Information 17 uses |
| relevant |
Words |
39 |
17 |
Attestation across the AI Supply Chain 9 uses |
| attribution |
Words |
38 |
13 |
N-gram search as posterior updating for data attribution 8 uses |
| best |
Words |
38 |
20 |
Attestation across the AI Supply Chain 8 uses |
| fact |
Words |
38 |
22 |
Will the New York Times Data Strike Have a Large Impact on ChatGPT? 5 uses |
| improve |
Words |
38 |
17 |
AI is driving the cost of polish down; some musings on fancy versus terse artifacts 6 uses |
| non |
Words |
38 |
16 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 9 uses |
| online |
Words |
38 |
18 |
The Paradox of Reuse in 2026: A Case of Quasi-Enclosure, or "Subsidized Club Goods that Sort of Look Like Public Goods" 5 uses |
| opt |
Words |
38 |
11 |
AI Artist or AI Art Thief? Innovation, Public Mandates, and the Case for Talking in Terms of Leverage 7 uses |
| start |
Words |
38 |
18 |
Tipping Points for Content Ecosystems 5 uses |
| able |
Words |
37 |
22 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 5 uses |
| api |
Words |
37 |
12 |
Coding agents are (1) a big deal, (2) very relevant to data leverage, and (3) able to help build tools that support data leverage! 9 uses |
| creation |
Words |
37 |
17 |
Building a Data Pipeworks for Democratic AI: From Human Knowledge to Records to AI Systems 9 uses |
| gpt |
Words |
37 |
11 |
Don’t give OpenAI all the credit for GPT-3: You might have helped create the latest “astonishing” advance in AI too 13 uses |
| language |
Words |
37 |
15 |
The Paradox of Reuse, Language Models Edition 10 uses |
| least |
Words |
37 |
22 |
Attestation across the AI Supply Chain 7 uses |
| paradigm |
Words |
37 |
23 |
The Coding Agent Data Deal 4 uses |
| attestations |
Words |
36 |
1 |
Attestation across the AI Supply Chain 36 uses |
| challenges |
Words |
36 |
17 |
Attestation across the AI Supply Chain 6 uses |
| clear |
Words |
36 |
16 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 7 uses |
| concerns |
Words |
36 |
19 |
Bing Rewards for the AI Age 3 uses |
| downstream |
Words |
36 |
14 |
Attestation across the AI Supply Chain 15 uses |
| evidence |
Words |
36 |
17 |
AI Dividends Without Taxing Compute, Automation, or Equity: A Presumptive Commons-Rent Tax Based on Capabilities and Data Dependence 12 uses |
| expect |
Words |
36 |
21 |
Public AI, Data Appraisal, and Data Debates 4 uses |
| free |
Words |
36 |
13 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 14 uses |
| perspective |
Words |
36 |
20 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 5 uses |
| privacy |
Words |
36 |
11 |
The Coding Agent Data Deal 19 uses |
| space |
Words |
36 |
18 |
Building a Data Pipeworks for Democratic AI: From Human Knowledge to Records to AI Systems 5 uses |
| whether |
Words |
36 |
22 |
Attestation across the AI Supply Chain 5 uses |
| write |
Words |
36 |
18 |
The WGA Strike is a Canary in the Coal Mine for AI Labor Concerns 3 uses |
| claim |
Words |
35 |
14 |
Attestation across the AI Supply Chain 12 uses |
| consumer |
Words |
35 |
12 |
Attestation across the AI Supply Chain 14 uses |
| creating |
Words |
35 |
21 |
Each Instance of "AI Utility" Stems from Some Human Act(s) of Information Recording and Ranking 4 uses |
| field |
Words |
35 |
20 |
On AI-driven Job Apocalypses and Collective Bargaining for Information 6 uses |
| focused |
Words |
35 |
20 |
AI is driving the cost of polish down; some musings on fancy versus terse artifacts 7 uses |
| github |
Words |
35 |
10 |
Bing Rewards for the AI Age 19 uses |
| industry |
Words |
35 |
18 |
Perplexity CEO's Interaction with Striking New York Times Workers Does Not Reflect Well on the AI Industry 6 uses |
| overall |
Words |
35 |
18 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 4 uses |
| platform |
Words |
35 |
16 |
Bing Rewards for the AI Age 6 uses |
| software |
Words |
35 |
15 |
Attestation across the AI Supply Chain 10 uses |
| stage |
Words |
35 |
6 |
Building a Data Pipeworks for Democratic AI: From Human Knowledge to Records to AI Systems 27 uses |
| true |
Words |
35 |
19 |
Is Zuckerberg right to say that your specific creative work has no value to AI? 5 uses |
| actual |
Words |
34 |
23 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 5 uses |
| continue |
Words |
34 |
18 |
Bing Rewards for the AI Age 6 uses |
| domain |
Words |
34 |
13 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 5 uses |
| enough |
Words |
34 |
16 |
Plural AI Data Alignment 5 uses |
| inputs |
Words |
34 |
10 |
Attestation across the AI Supply Chain 21 uses |
| news |
Words |
34 |
15 |
Will the New York Times Data Strike Have a Large Impact on ChatGPT? 8 uses |
| nyt |
Words |
34 |
6 |
Will the New York Times Data Strike Have a Large Impact on ChatGPT? 15 uses |
| training data |
2-grams |
124 |
33 |
Attestation across the AI Supply Chain 23 uses |
| ai systems |
2-grams |
93 |
33 |
Building a Data Pipeworks for Democratic AI: From Human Knowledge to Records to AI Systems 11 uses |
| generative ai |
2-grams |
67 |
17 |
Data Leverage Recap: December 2022 - April 2023 11 uses |
| public ai |
2-grams |
64 |
8 |
Public AI, Data Appraisal, and Data Debates 36 uses |
| data leverage |
2-grams |
63 |
18 |
Evaluation Data Leverage: Advances like "Deep Research" Highlight a Looming Opportunity for Bargaining Power 9 uses |
| collective action |
2-grams |
52 |
17 |
Algorithmic Collective Action With Two Collectives [crosspost] 17 uses |
| ai companies |
2-grams |
48 |
13 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 22 uses |
| data rules |
2-grams |
47 |
7 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 32 uses |
| ai products |
2-grams |
41 |
13 |
Attestation across the AI Supply Chain 14 uses |
| open source |
2-grams |
36 |
15 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 16 uses |
| coding agents |
2-grams |
35 |
5 |
The Coding Agent Data Deal 17 uses |
| collective bargaining |
2-grams |
35 |
15 |
On AI-driven Job Apocalypses and Collective Bargaining for Information 7 uses |
| ai models |
2-grams |
34 |
21 |
Building a Data Pipeworks for Democratic AI: From Human Knowledge to Records to AI Systems 3 uses |
| data strike |
2-grams |
34 |
6 |
Will the New York Times Data Strike Have a Large Impact on ChatGPT? 19 uses |
| data creators |
2-grams |
33 |
14 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 12 uses |
| blog post |
2-grams |
32 |
17 |
ChatGPT is Awesome and Scary: You Deserve Credit for the Good Parts (and Might Help Fix the Bad Parts) 8 uses |
| right now |
2-grams |
32 |
18 |
The Coding Agent Data Deal 4 uses |
| ai system |
2-grams |
28 |
12 |
Attestation across the AI Supply Chain 6 uses |
| data value |
2-grams |
28 |
10 |
Public AI, Data Appraisal, and Data Debates 8 uses |
| claude code |
2-grams |
27 |
5 |
The Coding Agent Data Deal 11 uses |
| data labor |
2-grams |
27 |
13 |
Don’t give OpenAI all the credit for GPT-3: You might have helped create the latest “astonishing” advance in AI too 5 uses |
| something like |
2-grams |
27 |
14 |
Attestation across the AI Supply Chain 6 uses |
| et al |
2-grams |
26 |
10 |
Attestation across the AI Supply Chain 6 uses |
| ai labs |
2-grams |
25 |
9 |
AI Labs Should Open Source Data Protection Technologies 9 uses |
| ai operators |
2-grams |
25 |
11 |
Reddit, StackOverflow, and Europe: All Trending Towards Data Dignity 6 uses |
| data flow |
2-grams |
25 |
12 |
Plural AI Data Alignment 4 uses |
| high quality |
2-grams |
25 |
11 |
The AI "Evaluation Crisis" Is an Opportunity to Get Data Flow Right 6 uses |
| knowledge work |
2-grams |
25 |
6 |
Evaluation Data Leverage: Advances like "Deep Research" Highlight a Looming Opportunity for Bargaining Power 8 uses |
| coding agent |
2-grams |
24 |
6 |
The Coding Agent Data Deal 17 uses |
| data protection |
2-grams |
24 |
5 |
AI Labs Should Open Source Data Protection Technologies 13 uses |
| model outputs |
2-grams |
24 |
7 |
Attestation across the AI Supply Chain 10 uses |
| open knowledge |
2-grams |
24 |
1 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 24 uses |
| model builders |
2-grams |
23 |
2 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 21 uses |
| new york |
2-grams |
22 |
8 |
A Harbinger of the Future of Content? The New York Times Starts a Data Strike 8 uses |
| york times |
2-grams |
22 |
8 |
A Harbinger of the Future of Content? The New York Times Starts a Data Strike 8 uses |
| data pipeworks |
2-grams |
21 |
2 |
Building a Data Pipeworks for Democratic AI: From Human Knowledge to Records to AI Systems 20 uses |
| data valuation |
2-grams |
21 |
8 |
[microblog] One book is worth "0.06%" benchmark points to AI; is "no different from noise". What gives? 5 uses |
| short term |
2-grams |
21 |
13 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 5 uses |
| ai progress |
2-grams |
20 |
9 |
AI is driving the cost of polish down; some musings on fancy versus terse artifacts 7 uses |
| evaluation data |
2-grams |
20 |
9 |
Evaluation Data Leverage: Advances like "Deep Research" Highlight a Looming Opportunity for Bargaining Power 6 uses |
| peer production |
2-grams |
20 |
9 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 8 uses |
| language models |
2-grams |
19 |
8 |
The Paradox of Reuse, Language Models Edition 8 uses |
| model weights |
2-grams |
19 |
14 |
AI progress as quasi-public good production 4 uses |
| public goods |
2-grams |
19 |
7 |
AI progress as quasi-public good production 5 uses |
| ai capabilities |
2-grams |
18 |
11 |
Building a Data Pipeworks for Democratic AI: From Human Knowledge to Records to AI Systems 5 uses |
| common crawl |
2-grams |
18 |
7 |
ChatGPT is Awesome and Scary: You Deserve Credit for the Good Parts (and Might Help Fix the Bad Parts) 7 uses |
| data strikes |
2-grams |
18 |
8 |
Will the New York Times Data Strike Have a Large Impact on ChatGPT? 6 uses |
| decision making |
2-grams |
18 |
12 |
Attestation across the AI Supply Chain 5 uses |
| look like |
2-grams |
18 |
11 |
Evaluation Data Leverage: Advances like "Deep Research" Highlight a Looming Opportunity for Bargaining Power 3 uses |
| ai data |
2-grams |
17 |
11 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 5 uses |
| data markets |
2-grams |
17 |
7 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 5 uses |
| machine learning |
2-grams |
17 |
7 |
Building a Data Pipeworks for Democratic AI: From Human Knowledge to Records to AI Systems 8 uses |
| search engines |
2-grams |
17 |
9 |
The Paradox of Reuse, Language Models Edition 6 uses |
| ai developers |
2-grams |
16 |
8 |
Attestation across the AI Supply Chain 8 uses |
| ai outputs |
2-grams |
16 |
6 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 5 uses |
| data transparency |
2-grams |
16 |
8 |
Two natural allies of a "Data Transparency" agenda: capabilities forecasters and social simulators 5 uses |
| deep research |
2-grams |
16 |
3 |
Evaluation Data Leverage: Advances like "Deep Research" Highlight a Looming Opportunity for Bargaining Power 14 uses |
| frontier ai |
2-grams |
16 |
6 |
AI Dividends Without Taxing Compute, Automation, or Equity: A Presumptive Commons-Rent Tax Based on Capabilities and Data Dependence 5 uses |
| going forward |
2-grams |
16 |
12 |
AI Labs Should Open Source Data Protection Technologies 3 uses |
| long run |
2-grams |
16 |
10 |
The WGA Strike is a Canary in the Coal Mine for AI Labor Concerns 3 uses |
| supply chain |
2-grams |
16 |
4 |
Attestation across the AI Supply Chain 12 uses |
| wikimedia commons |
2-grams |
16 |
14 |
Public AI, Data Appraisal, and Data Debates 2 uses |
| bargaining power |
2-grams |
15 |
12 |
Attestation across the AI Supply Chain 3 uses |
| data contributions |
2-grams |
15 |
8 |
Plural AI Data Alignment 4 uses |
| data dividends |
2-grams |
15 |
5 |
AI Dividends Without Taxing Compute, Automation, or Equity: A Presumptive Commons-Rent Tax Based on Capabilities and Data Dependence 11 uses |
| looks like |
2-grams |
15 |
10 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 4 uses |
| model builder |
2-grams |
15 |
1 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 15 uses |
| post training |
2-grams |
15 |
10 |
Two natural allies of a "Data Transparency" agenda: capabilities forecasters and social simulators 3 uses |
| tipping points |
2-grams |
15 |
5 |
Tipping Points for Content Ecosystems 9 uses |
| data attribution |
2-grams |
14 |
6 |
N-gram search as posterior updating for data attribution 7 uses |
| data dividend |
2-grams |
14 |
6 |
AI Dividends Without Taxing Compute, Automation, or Equity: A Presumptive Commons-Rent Tax Based on Capabilities and Data Dependence 7 uses |
| data sharing |
2-grams |
14 |
7 |
Public AI, Data Appraisal, and Data Debates 6 uses |
| public good |
2-grams |
14 |
3 |
AI progress as quasi-public good production 10 uses |
| stack exchange |
2-grams |
14 |
5 |
Tipping Points for Content Ecosystems 6 uses |
| use ai |
2-grams |
14 |
7 |
How do we know our AI output is good? Double checks, bar charts, vibes, and training data. 3 uses |
| ai field |
2-grams |
13 |
8 |
On AI-driven Job Apocalypses and Collective Bargaining for Information 4 uses |
| data contribution |
2-grams |
13 |
10 |
Coding agents are (1) a big deal, (2) very relevant to data leverage, and (3) able to help build tools that support data leverage! 2 uses |
| data dependent |
2-grams |
13 |
9 |
ChatGPT is Awesome and Scary: You Deserve Credit for the Good Parts (and Might Help Fix the Bad Parts) 2 uses |
| data related |
2-grams |
13 |
8 |
Building a Data Pipeworks for Democratic AI: From Human Knowledge to Records to AI Systems 3 uses |
| high level |
2-grams |
13 |
10 |
Two natural allies of a "Data Transparency" agenda: capabilities forecasters and social simulators 3 uses |
| like wikipedia |
2-grams |
13 |
6 |
Bing Rewards for the AI Age 4 uses |
| long term |
2-grams |
13 |
10 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 4 uses |
| new ai |
2-grams |
13 |
11 |
Bing Rewards for the AI Age 3 uses |
| platforms like |
2-grams |
13 |
5 |
Bing Rewards for the AI Age 4 uses |
| pre training |
2-grams |
13 |
10 |
Selling AGI like AG1: Will Consumers Push Back Against Proprietary Blends of Herbs and of Data? 2 uses |
| ai generated |
2-grams |
12 |
6 |
Attestation across the AI Supply Chain 4 uses |
| ai industry |
2-grams |
12 |
7 |
Perplexity CEO's Interaction with Striking New York Times Workers Does Not Reflect Well on the AI Industry 6 uses |
| ai lab |
2-grams |
12 |
7 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 4 uses |
| ai product |
2-grams |
12 |
5 |
Attestation across the AI Supply Chain 7 uses |
| ai supply |
2-grams |
12 |
4 |
Attestation across the AI Supply Chain 7 uses |
| data sources |
2-grams |
12 |
9 |
ChatGPT is Awesome and Scary: You Deserve Credit for the Good Parts (and Might Help Fix the Bad Parts) 3 uses |
| dataset documentation |
2-grams |
12 |
6 |
How do we know our AI output is good? Double checks, bar charts, vibes, and training data. 5 uses |
| human knowledge |
2-grams |
12 |
7 |
Building a Data Pipeworks for Democratic AI: From Human Knowledge to Records to AI Systems 3 uses |
| language model |
2-grams |
12 |
9 |
The Paradox of Reuse, Language Models Edition 2 uses |
| large scale |
2-grams |
12 |
7 |
Building a Data Pipeworks for Democratic AI: From Human Knowledge to Records to AI Systems 3 uses |
| model training |
2-grams |
12 |
7 |
Building a Data Pipeworks for Democratic AI: From Human Knowledge to Records to AI Systems 4 uses |
| powerful ai |
2-grams |
12 |
4 |
Tipping Points for Content Ecosystems 7 uses |
| social dilemmas |
2-grams |
12 |
3 |
Building a Data Pipeworks for Democratic AI: From Human Knowledge to Records to AI Systems 10 uses |
| ai tools |
2-grams |
11 |
7 |
The Paradox of Reuse in 2026: A Case of Quasi-Enclosure, or "Subsidized Club Goods that Sort of Look Like Public Goods" 4 uses |
| ai training |
2-grams |
11 |
7 |
A Harbinger of the Future of Content? The New York Times Starts a Data Strike 3 uses |
| content ecosystems |
2-grams |
11 |
5 |
Tipping Points for Content Ecosystems 4 uses |
| data flows |
2-grams |
11 |
6 |
Plural AI Data Alignment 6 uses |
| feedback loops |
2-grams |
11 |
9 |
Building a Data Pipeworks for Democratic AI: From Human Knowledge to Records to AI Systems 2 uses |
| knowledge production |
2-grams |
11 |
2 |
The Paradox of Reuse in 2026: A Case of Quasi-Enclosure, or "Subsidized Club Goods that Sort of Look Like Public Goods" 10 uses |
| model version |
2-grams |
11 |
1 |
Attestation across the AI Supply Chain 11 uses |
| one way |
2-grams |
11 |
9 |
AI is driving the cost of polish down; some musings on fancy versus terse artifacts 2 uses |
| online platforms |
2-grams |
11 |
8 |
Data Leverage Recap: December 2022 - April 2023 3 uses |
| supervised learning |
2-grams |
11 |
2 |
The AI "Evaluation Crisis" Is an Opportunity to Get Data Flow Right 10 uses |
| synthetic data |
2-grams |
11 |
7 |
On AI-driven Job Apocalypses and Collective Bargaining for Information 2 uses |
| tech companies |
2-grams |
11 |
8 |
Don’t give OpenAI all the credit for GPT-3: You might have helped create the latest “astonishing” advance in AI too 2 uses |
| training set |
2-grams |
11 |
10 |
Plural AI Data Alignment 2 uses |
| wikimedia enterprise |
2-grams |
11 |
5 |
AI Dividends Without Taxing Compute, Automation, or Equity: A Presumptive Commons-Rent Tax Based on Capabilities and Data Dependence 4 uses |
| ai evaluation |
2-grams |
10 |
3 |
Attestation across the AI Supply Chain 4 uses |
| ai model |
2-grams |
10 |
6 |
Live by the free-content-for-training sword, die by the free-content-for-training sword 2 uses |
| ai technologies |
2-grams |
10 |
7 |
Data Leverage Recap: December 2022 - April 2023 3 uses |
| attestation objects |
2-grams |
10 |
2 |
Attestation across the AI Supply Chain 9 uses |
| data centric |
2-grams |
10 |
9 |
Two natural allies of a "Data Transparency" agenda: capabilities forecasters and social simulators 2 uses |
| data dependence |
2-grams |
10 |
1 |
AI Dividends Without Taxing Compute, Automation, or Equity: A Presumptive Commons-Rent Tax Based on Capabilities and Data Dependence 10 uses |
| different groups |
2-grams |
10 |
5 |
Plural AI Data Alignment 3 uses |
| evaluation crisis |
2-grams |
10 |
2 |
The AI "Evaluation Crisis" Is an Opportunity to Get Data Flow Right 6 uses |
| lot like |
2-grams |
10 |
7 |
The AI "Evaluation Crisis" Is an Opportunity to Get Data Flow Right 3 uses |
| many people |
2-grams |
10 |
6 |
Attestation across the AI Supply Chain 3 uses |
| new data |
2-grams |
10 |
5 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 5 uses |
| public interest |
2-grams |
10 |
6 |
Public AI, Data Appraisal, and Data Debates 3 uses |
| quasi public |
2-grams |
10 |
2 |
AI progress as quasi-public good production 9 uses |
| reddit data |
2-grams |
10 |
4 |
Public AI, Data Appraisal, and Data Debates 6 uses |
| search engine |
2-grams |
10 |
8 |
Bing Rewards for the AI Age 3 uses |
| systems like |
2-grams |
10 |
6 |
Bing Rewards for the AI Age 3 uses |
| wikipedia article |
2-grams |
10 |
7 |
Google and TikTok rank bundles of information; ChatGPT ranks grains. 2 uses |
| ai governance |
2-grams |
9 |
6 |
AI Technologies are System Maps, and You are a Cartographer 2 uses |
| ai output |
2-grams |
9 |
5 |
How do we know our AI output is good? Double checks, bar charts, vibes, and training data. 3 uses |
| another way |
2-grams |
9 |
7 |
Attestation across the AI Supply Chain 3 uses |
| certain types |
2-grams |
9 |
5 |
The Paradox of Reuse in 2026: A Case of Quasi-Enclosure, or "Subsidized Club Goods that Sort of Look Like Public Goods" 3 uses |
| clear data |
2-grams |
9 |
4 |
AI Dividends Without Taxing Compute, Automation, or Equity: A Presumptive Commons-Rent Tax Based on Capabilities and Data Dependence 4 uses |
| club goods |
2-grams |
9 |
2 |
The Paradox of Reuse in 2026: A Case of Quasi-Enclosure, or "Subsidized Club Goods that Sort of Look Like Public Goods" 8 uses |
| data agency |
2-grams |
9 |
5 |
Plural AI Data Alignment 3 uses |
| data appraisal |
2-grams |
9 |
2 |
Public AI, Data Appraisal, and Data Debates 7 uses |
| data deals |
2-grams |
9 |
4 |
Public AI, Data Appraisal, and Data Debates 5 uses |
| high stakes |
2-grams |
9 |
6 |
How collective bargaining for information, public AI, and HCI research all fit together 2 uses |
| intellectual property |
2-grams |
9 |
7 |
A Harbinger of the Future of Content? The New York Times Starts a Data Strike 3 uses |
| like chatgpt |
2-grams |
9 |
6 |
Data Leverage Recap: December 2022 - April 2023 3 uses |
| many ways |
2-grams |
9 |
6 |
Google and TikTok rank bundles of information; ChatGPT ranks grains. 2 uses |
| ongoing discussions |
2-grams |
9 |
7 |
Data Leverage Recap: December 2022 - April 2023 2 uses |
| power concentration |
2-grams |
9 |
5 |
On AI-driven Job Apocalypses and Collective Bargaining for Information 5 uses |
| quasi enclosure |
2-grams |
9 |
3 |
The Paradox of Reuse in 2026: A Case of Quasi-Enclosure, or "Subsidized Club Goods that Sort of Look Like Public Goods" 5 uses |
| reinforcement learning |
2-grams |
9 |
8 |
Each Instance of "AI Utility" Stems from Some Human Act(s) of Information Recording and Ranking 2 uses |
| social simulation |
2-grams |
9 |
3 |
Two natural allies of a "Data Transparency" agenda: capabilities forecasters and social simulators 7 uses |
| source software |
2-grams |
9 |
4 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 6 uses |
| training sets |
2-grams |
9 |
6 |
AI Artist or AI Art Thief? Innovation, Public Mandates, and the Case for Talking in Terms of Leverage 4 uses |
| white collar |
2-grams |
9 |
3 |
On AI-driven Job Apocalypses and Collective Bargaining for Information 5 uses |
| ai auditing |
2-grams |
8 |
2 |
AI Dividends Without Taxing Compute, Automation, or Equity: A Presumptive Commons-Rent Tax Based on Capabilities and Data Dependence 5 uses |
| ai related |
2-grams |
8 |
6 |
AI Dividends Without Taxing Compute, Automation, or Equity: A Presumptive Commons-Rent Tax Based on Capabilities and Data Dependence 3 uses |
| bullet points |
2-grams |
8 |
2 |
AI is driving the cost of polish down; some musings on fancy versus terse artifacts 7 uses |
| data collectives |
2-grams |
8 |
3 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 5 uses |
| data provenance |
2-grams |
8 |
5 |
AI Dividends Without Taxing Compute, Automation, or Equity: A Presumptive Commons-Rent Tax Based on Capabilities and Data Dependence 4 uses |
| data values |
2-grams |
8 |
4 |
N-gram search as posterior updating for data attribution 3 uses |
| eval data |
2-grams |
8 |
5 |
Evaluation Data Leverage: Advances like "Deep Research" Highlight a Looming Opportunity for Bargaining Power 4 uses |
| evaluation labor |
2-grams |
8 |
2 |
Attestation across the AI Supply Chain 7 uses |
| high value |
2-grams |
8 |
4 |
The Coding Agent Data Deal 3 uses |
| key points |
2-grams |
8 |
3 |
Data Leverage Recap: December 2022 - April 2023 6 uses |
| large number |
2-grams |
8 |
7 |
Bing Rewards for the AI Age 2 uses |
| many cases |
2-grams |
8 |
6 |
Public AI, Data Appraisal, and Data Debates 2 uses |
| many models |
2-grams |
8 |
2 |
"Many Models" and "Track Changes" for AI: Some Thoughts on LLM Interfaces 7 uses |
| model card |
2-grams |
8 |
2 |
ChatGPT is Awesome and Scary: You Deserve Credit for the Good Parts (and Might Help Fix the Bad Parts) 7 uses |
| near term |
2-grams |
8 |
7 |
AI Labs Should Open Source Data Protection Technologies 2 uses |
| previous post |
2-grams |
8 |
5 |
Will the New York Times Data Strike Have a Large Impact on ChatGPT? 2 uses |
| social media |
2-grams |
8 |
8 |
AI Technologies are System Maps, and You are a Cartographer 1 use |
| stack overflow |
2-grams |
8 |
6 |
Tipping Points for Content Ecosystems 2 uses |
| value estimation |
2-grams |
8 |
3 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 6 uses |
| washington post |
2-grams |
8 |
4 |
Perplexity CEO's Interaction with Striking New York Times Workers Does Not Reflect Well on the AI Industry 3 uses |
| wealth fund |
2-grams |
8 |
3 |
AI Dividends Without Taxing Compute, Automation, or Equity: A Presumptive Commons-Rent Tax Based on Capabilities and Data Dependence 6 uses |
| agent based |
2-grams |
7 |
3 |
Building a Data Pipeworks for Democratic AI: From Human Knowledge to Records to AI Systems 4 uses |
| ai company |
2-grams |
7 |
4 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 3 uses |
| ai operator |
2-grams |
7 |
7 |
Bing Rewards for the AI Age 1 use |
| ai research |
2-grams |
7 |
6 |
On AI-driven Job Apocalypses and Collective Bargaining for Information 2 uses |
| ai safety |
2-grams |
7 |
7 |
Plural AI Data Alignment 1 use |
| anti scraping |
2-grams |
7 |
2 |
Attestation across the AI Supply Chain 6 uses |
| around data |
2-grams |
7 |
7 |
AI Technologies are System Maps, and You are a Cartographer 1 use |
| bing chat |
2-grams |
7 |
2 |
Bing Rewards for the AI Age 6 uses |
| commons data |
2-grams |
7 |
2 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 4 uses |
| data commons |
2-grams |
7 |
3 |
AI Dividends Without Taxing Compute, Automation, or Equity: A Presumptive Commons-Rent Tax Based on Capabilities and Data Dependence 4 uses |
| data control |
2-grams |
7 |
2 |
The Coding Agent Data Deal 5 uses |
| data dignity |
2-grams |
7 |
5 |
Reddit, StackOverflow, and Europe: All Trending Towards Data Dignity 3 uses |
| data licensing |
2-grams |
7 |
4 |
Public AI, Data Appraisal, and Data Debates 2 uses |
| data poisoning |
2-grams |
7 |
4 |
A Short Guide to Data Strikes and Conscious Data Contribution in the Context of 2026 Frontier AI 4 uses |
| data records |
2-grams |
7 |
4 |
Building a Data Pipeworks for Democratic AI: From Human Knowledge to Records to AI Systems 3 uses |
| data workers |
2-grams |
7 |
3 |
Attestation across the AI Supply Chain 5 uses |
| digital commons |
2-grams |
7 |
5 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 3 uses |
| economic impacts |
2-grams |
7 |
5 |
On AI-driven Job Apocalypses and Collective Bargaining for Information 2 uses |
| generated content |
2-grams |
7 |
5 |
Reddit, StackOverflow, and Europe: All Trending Towards Data Dignity 2 uses |
| improve ai |
2-grams |
7 |
5 |
The Coding Agent Data Deal 3 uses |
| knowledge workers |
2-grams |
7 |
3 |
Evaluation Data Leverage: Advances like "Deep Research" Highlight a Looming Opportunity for Bargaining Power 4 uses |
| large language |
2-grams |
7 |
4 |
ChatGPT is Awesome and Scary: You Deserve Credit for the Good Parts (and Might Help Fix the Bad Parts) 3 uses |
| leave one |
2-grams |
7 |
4 |
Is Zuckerberg right to say that your specific creative work has no value to AI? 3 uses |
| legal action |
2-grams |
7 |
2 |
A Harbinger of the Future of Content? The New York Times Starts a Data Strike 6 uses |
| level ai |
2-grams |
7 |
1 |
Tipping Points for Content Ecosystems 7 uses |
| llm capabilities |
2-grams |
7 |
2 |
Bing Rewards for the AI Age 4 uses |
| llm interfaces |
2-grams |
7 |
2 |
"Many Models" and "Track Changes" for AI: Some Thoughts on LLM Interfaces 6 uses |
| model stealing |
2-grams |
7 |
1 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 7 uses |
| napkin math |
2-grams |
7 |
5 |
Tipping Points for Content Ecosystems 2 uses |
| peer review |
2-grams |
7 |
3 |
AI is driving the cost of polish down; some musings on fancy versus terse artifacts 4 uses |
| public domain |
2-grams |
7 |
5 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 2 uses |
| reddit users |
2-grams |
7 |
6 |
Public AI, Data Appraisal, and Data Debates 2 uses |
| robots txt |
2-grams |
7 |
3 |
A Harbinger of the Future of Content? The New York Times Starts a Data Strike 5 uses |
| third party |
2-grams |
7 |
4 |
Attestation across the AI Supply Chain 3 uses |
| towards data |
2-grams |
7 |
6 |
Reddit, StackOverflow, and Europe: All Trending Towards Data Dignity 2 uses |
| win win |
2-grams |
7 |
3 |
Attestation across the AI Supply Chain 4 uses |
| ai age |
2-grams |
6 |
6 |
Bing Rewards for the AI Age 1 use |
| ai art |
2-grams |
6 |
3 |
AI Artist or AI Art Thief? Innovation, Public Mandates, and the Case for Talking in Terms of Leverage 4 uses |
| ai assisted |
2-grams |
6 |
4 |
"People First" Policy Ideas that Complement Each Other (through better data flow) 2 uses |
| ai impact |
2-grams |
6 |
2 |
The Paradox of Reuse in 2026: A Case of Quasi-Enclosure, or "Subsidized Club Goods that Sort of Look Like Public Goods" 5 uses |
| ai network |
2-grams |
6 |
2 |
Public AI, Data Appraisal, and Data Debates 5 uses |
| ai use |
2-grams |
6 |
5 |
The WGA Strike is a Canary in the Coal Mine for AI Labor Concerns 2 uses |
| ai users |
2-grams |
6 |
3 |
Attestation across the AI Supply Chain 4 uses |
| algorithmic collective |
2-grams |
6 |
3 |
Algorithmic Collective Action With Two Collectives [crosspost] 4 uses |
| api keys |
2-grams |
6 |
3 |
Coding agents are (1) a big deal, (2) very relevant to data leverage, and (3) able to help build tools that support data leverage! 4 uses |
| artificial intelligence |
2-grams |
6 |
4 |
The WGA Strike is a Canary in the Coal Mine for AI Labor Concerns 2 uses |
| bar charts |
2-grams |
6 |
1 |
How do we know our AI output is good? Double checks, bar charts, vibes, and training data. 6 uses |
| big deal |
2-grams |
6 |
4 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 2 uses |
| cartographic labor |
2-grams |
6 |
3 |
Data Leverage Recap: December 2022 - April 2023 3 uses |
| commons rent |
2-grams |
6 |
1 |
AI Dividends Without Taxing Compute, Automation, or Equity: A Presumptive Commons-Rent Tax Based on Capabilities and Data Dependence 6 uses |
| computer interaction |
2-grams |
6 |
3 |
Building a Data Pipeworks for Democratic AI: From Human Knowledge to Records to AI Systems 4 uses |
| content creators |
2-grams |
6 |
4 |
AI Labs Should Open Source Data Protection Technologies 2 uses |
| contract templates |
2-grams |
6 |
1 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 6 uses |
| current paradigm |
2-grams |
6 |
6 |
AI Labs Should Open Source Data Protection Technologies 1 use |
| current situation |
2-grams |
6 |
4 |
The Paradox of Reuse in 2026: A Case of Quasi-Enclosure, or "Subsidized Club Goods that Sort of Look Like Public Goods" 3 uses |
| data contributors |
2-grams |
6 |
4 |
Plural AI Data Alignment 3 uses |
| data creation |
2-grams |
6 |
5 |
The AI "Evaluation Crisis" Is an Opportunity to Get Data Flow Right 2 uses |
| data generating |
2-grams |
6 |
3 |
Evaluation Data Leverage: Advances like "Deep Research" Highlight a Looming Opportunity for Bargaining Power 3 uses |
| data points |
2-grams |
6 |
3 |
AI Dividends Without Taxing Compute, Automation, or Equity: A Presumptive Commons-Rent Tax Based on Capabilities and Data Dependence 3 uses |
| data policy |
2-grams |
6 |
3 |
The Coding Agent Data Deal 3 uses |
| data sellers |
2-grams |
6 |
2 |
Public AI, Data Appraisal, and Data Debates 3 uses |
| data use |
2-grams |
6 |
3 |
The Coding Agent Data Deal 3 uses |
| double check |
2-grams |
6 |
2 |
How do we know our AI output is good? Double checks, bar charts, vibes, and training data. 5 uses |
| downstream use |
2-grams |
6 |
2 |
Attestation across the AI Supply Chain 4 uses |
| efforts like |
2-grams |
6 |
5 |
The AI "Evaluation Crisis" Is an Opportunity to Get Data Flow Right 2 uses |
| explained data |
2-grams |
6 |
1 |
AI Dividends Without Taxing Compute, Automation, or Equity: A Presumptive Commons-Rent Tax Based on Capabilities and Data Dependence 6 uses |
| fancy artifacts |
2-grams |
6 |
1 |
AI is driving the cost of polish down; some musings on fancy versus terse artifacts 6 uses |
| foundation model |
2-grams |
6 |
3 |
AI Dividends Without Taxing Compute, Automation, or Equity: A Presumptive Commons-Rent Tax Based on Capabilities and Data Dependence 4 uses |
| getting better |
2-grams |
6 |
2 |
The Paradox of Reuse in 2026: A Case of Quasi-Enclosure, or "Subsidized Club Goods that Sort of Look Like Public Goods" 5 uses |
| good faith |
2-grams |
6 |
4 |
Bing Rewards for the AI Age 2 uses |
| gram search |
2-grams |
6 |
1 |
N-gram search as posterior updating for data attribution 6 uses |
| human computer |
2-grams |
6 |
3 |
Building a Data Pipeworks for Democratic AI: From Human Knowledge to Records to AI Systems 4 uses |
| human factors |
2-grams |
6 |
1 |
Building a Data Pipeworks for Democratic AI: From Human Knowledge to Records to AI Systems 6 uses |
| human feedback |
2-grams |
6 |
4 |
ChatGPT is Awesome and Scary: You Deserve Credit for the Good Parts (and Might Help Fix the Bad Parts) 3 uses |
| key idea |
2-grams |
6 |
6 |
Plural AI Data Alignment 1 use |
| knowledge commons |
2-grams |
6 |
4 |
The Paradox of Reuse in 2026: A Case of Quasi-Enclosure, or "Subsidized Club Goods that Sort of Look Like Public Goods" 3 uses |
| model capabilities |
2-grams |
6 |
6 |
Data Leverage Recap: December 2022 - April 2023 1 use |
| model performance |
2-grams |
6 |
3 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 4 uses |
| news organizations |
2-grams |
6 |
3 |
Will the New York Times Data Strike Have a Large Impact on ChatGPT? 4 uses |
| nyt content |
2-grams |
6 |
2 |
Will the New York Times Data Strike Have a Large Impact on ChatGPT? 3 uses |
| open data |
2-grams |
6 |
3 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 3 uses |
| open web |
2-grams |
6 |
4 |
Tipping Points for Content Ecosystems 2 uses |
| organizations like |
2-grams |
6 |
5 |
Public AI, Data Appraisal, and Data Debates 2 uses |
| output provenance |
2-grams |
6 |
1 |
Attestation across the AI Supply Chain 6 uses |
| pipeworks model |
2-grams |
6 |
2 |
Building a Data Pipeworks for Democratic AI: From Human Knowledge to Records to AI Systems 5 uses |
| playing field |
2-grams |
6 |
5 |
Public AI, Data Appraisal, and Data Debates 2 uses |
| please let |
2-grams |
6 |
6 |
AI Technologies are System Maps, and You are a Cartographer 1 use |
| points hold |
2-grams |
6 |
1 |
Data Leverage Recap: December 2022 - April 2023 6 uses |
| post want |
2-grams |
6 |
6 |
Reddit, StackOverflow, and Europe: All Trending Towards Data Dignity 1 use |
| provenance initiative |
2-grams |
6 |
4 |
AI Dividends Without Taxing Compute, Automation, or Equity: A Presumptive Commons-Rent Tax Based on Capabilities and Data Dependence 3 uses |
| public bodies |
2-grams |
6 |
4 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 2 uses |
| public ownership |
2-grams |
6 |
1 |
AI Dividends Without Taxing Compute, Automation, or Equity: A Presumptive Commons-Rent Tax Based on Capabilities and Data Dependence 6 uses |
| put another |
2-grams |
6 |
4 |
Attestation across the AI Supply Chain 3 uses |
| put simply |
2-grams |
6 |
6 |
The Paradox of Reuse, Language Models Edition 1 use |
| raw data |
2-grams |
6 |
3 |
AI is driving the cost of polish down; some musings on fancy versus terse artifacts 3 uses |
| really good |
2-grams |
6 |
6 |
Selling AGI like AG1: Will Consumers Push Back Against Proprietary Blends of Herbs and of Data? 1 use |
| record creation |
2-grams |
6 |
2 |
Building a Data Pipeworks for Democratic AI: From Human Knowledge to Records to AI Systems 5 uses |
| sanders ai |
2-grams |
6 |
1 |
AI Dividends Without Taxing Compute, Automation, or Equity: A Presumptive Commons-Rent Tax Based on Capabilities and Data Dependence 6 uses |
| seaside village |
2-grams |
6 |
1 |
AI Technologies are System Maps, and You are a Cartographer 6 uses |
| seems like |
2-grams |
6 |
5 |
The Paradox of Reuse in 2026: A Case of Quasi-Enclosure, or "Subsidized Club Goods that Sort of Look Like Public Goods" 2 uses |
| small group |
2-grams |
6 |
4 |
Reddit, StackOverflow, and Europe: All Trending Towards Data Dignity 3 uses |
| small set |
2-grams |
6 |
3 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 3 uses |
| source ai |
2-grams |
6 |
3 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 3 uses |
| status quo |
2-grams |
6 |
4 |
Reddit, StackOverflow, and Europe: All Trending Towards Data Dignity 2 uses |
| structured data |
2-grams |
6 |
4 |
The AI "Evaluation Crisis" Is an Opportunity to Get Data Flow Right 3 uses |
| tech industry |
2-grams |
6 |
4 |
Live by the free-content-for-training sword, die by the free-content-for-training sword 2 uses |
| test set |
2-grams |
6 |
3 |
Plural AI Data Alignment 3 uses |
| thought experiment |
2-grams |
6 |
3 |
AI Technologies are System Maps, and You are a Cartographer 4 uses |
| thus far |
2-grams |
6 |
5 |
The AI "Evaluation Crisis" Is an Opportunity to Get Data Flow Right 2 uses |
| train ai |
2-grams |
6 |
6 |
The WGA Strike is a Canary in the Coal Mine for AI Labor Concerns 1 use |
| training dataset |
2-grams |
6 |
5 |
A Harbinger of the Future of Content? The New York Times Starts a Data Strike 2 uses |
| upstream data |
2-grams |
6 |
5 |
Attestation across the AI Supply Chain 2 uses |
| using ai |
2-grams |
6 |
5 |
The WGA Strike is a Canary in the Coal Mine for AI Labor Concerns 2 uses |
| web app |
2-grams |
6 |
4 |
The Coding Agent Data Deal 3 uses |
| web search |
2-grams |
6 |
3 |
The Coding Agent Data Deal 3 uses |
| wide variety |
2-grams |
6 |
4 |
The WGA Strike is a Canary in the Coal Mine for AI Labor Concerns 2 uses |
| ablation experiments |
2-grams |
5 |
4 |
Public AI, Data Appraisal, and Data Debates 2 uses |
| action problem |
2-grams |
5 |
4 |
A Harbinger of the Future of Content? The New York Times Starts a Data Strike 2 uses |
| agent paradigm |
2-grams |
5 |
2 |
The Coding Agent Data Deal 4 uses |
| agent transcripts |
2-grams |
5 |
3 |
Coding agents are (1) a big deal, (2) very relevant to data leverage, and (3) able to help build tools that support data leverage! 2 uses |
| ai agents |
2-grams |
5 |
3 |
Each Instance of "AI Utility" Stems from Some Human Act(s) of Information Recording and Ranking 2 uses |
| ai alignment |
2-grams |
5 |
3 |
Plural AI Data Alignment 2 uses |
| ai body |
2-grams |
5 |
1 |
Public AI, Data Appraisal, and Data Debates 5 uses |
| ai dividend |
2-grams |
5 |
1 |
AI Dividends Without Taxing Compute, Automation, or Equity: A Presumptive Commons-Rent Tax Based on Capabilities and Data Dependence 5 uses |
| ai social |
2-grams |
5 |
2 |
Two natural allies of a "Data Transparency" agenda: capabilities forecasters and social simulators 3 uses |
| ai utility |
2-grams |
5 |
3 |
Each Instance of "AI Utility" Stems from Some Human Act(s) of Information Recording and Ranking 2 uses |
| almost always |
2-grams |
5 |
4 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 2 uses |
| almost certainly |
2-grams |
5 |
5 |
Don’t give OpenAI all the credit for GPT-3: You might have helped create the latest “astonishing” advance in AI too 1 use |
| benchmark contamination |
2-grams |
5 |
3 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 3 uses |
| new york times |
3-grams |
22 |
8 |
A Harbinger of the Future of Content? The New York Times Starts a Data Strike 8 uses |
| generative ai systems |
3-grams |
12 |
4 |
Bing Rewards for the AI Age 6 uses |
| ai supply chain |
3-grams |
10 |
3 |
Attestation across the AI Supply Chain 7 uses |
| clear data rules |
3-grams |
9 |
4 |
AI Dividends Without Taxing Compute, Automation, or Equity: A Presumptive Commons-Rent Tax Based on Capabilities and Data Dependence 4 uses |
| evaluation data leverage |
3-grams |
9 |
2 |
Evaluation Data Leverage: Advances like "Deep Research" Highlight a Looming Opportunity for Bargaining Power 6 uses |
| open source software |
3-grams |
9 |
4 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 6 uses |
| frontier ai auditing |
3-grams |
7 |
2 |
AI Dividends Without Taxing Compute, Automation, or Equity: A Presumptive Commons-Rent Tax Based on Capabilities and Data Dependence 4 uses |
| pre training data |
3-grams |
7 |
7 |
Bing Rewards for the AI Age 1 use |
| algorithmic collective action |
3-grams |
6 |
3 |
Algorithmic Collective Action With Two Collectives [crosspost] 4 uses |
| data provenance initiative |
3-grams |
6 |
4 |
AI Dividends Without Taxing Compute, Automation, or Equity: A Presumptive Commons-Rent Tax Based on Capabilities and Data Dependence 3 uses |
| human computer interaction |
3-grams |
6 |
3 |
Building a Data Pipeworks for Democratic AI: From Human Knowledge to Records to AI Systems 4 uses |
| key points hold |
3-grams |
6 |
1 |
Data Leverage Recap: December 2022 - April 2023 6 uses |
| public ai network |
3-grams |
6 |
2 |
Public AI, Data Appraisal, and Data Debates 5 uses |
| put another way |
3-grams |
6 |
4 |
Attestation across the AI Supply Chain 3 uses |
| quasi public good |
3-grams |
6 |
1 |
AI progress as quasi-public good production 6 uses |
| ai evaluation crisis |
3-grams |
5 |
2 |
The AI "Evaluation Crisis" Is an Opportunity to Get Data Flow Right 3 uses |
| collective action problem |
3-grams |
5 |
4 |
A Harbinger of the Future of Content? The New York Times Starts a Data Strike 2 uses |
| conscious data contribution |
3-grams |
5 |
4 |
A Short Guide to Data Strikes and Conscious Data Contribution in the Context of 2026 Frontier AI 2 uses |
| data dependent systems |
3-grams |
5 |
4 |
Building a Data Pipeworks for Democratic AI: From Human Knowledge to Records to AI Systems 2 uses |
| foundation model transparency |
3-grams |
5 |
2 |
AI Dividends Without Taxing Compute, Automation, or Equity: A Presumptive Commons-Rent Tax Based on Capabilities and Data Dependence 4 uses |
| large language models |
3-grams |
5 |
3 |
ChatGPT is Awesome and Scary: You Deserve Credit for the Good Parts (and Might Help Fix the Bad Parts) 3 uses |
| open source ai |
3-grams |
5 |
2 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 3 uses |
| platforms like wikipedia |
3-grams |
5 |
3 |
Bing Rewards for the AI Age 2 uses |
| presumptive commons rent |
3-grams |
5 |
1 |
AI Dividends Without Taxing Compute, Automation, or Equity: A Presumptive Commons-Rent Tax Based on Capabilities and Data Dependence 5 uses |
| public ai body |
3-grams |
5 |
1 |
Public AI, Data Appraisal, and Data Debates 5 uses |
| public ai labs |
3-grams |
5 |
2 |
Public AI, Data Appraisal, and Data Debates 4 uses |
| training data attribution |
3-grams |
5 |
4 |
N-gram search as posterior updating for data attribution 2 uses |
| using coding agents |
3-grams |
5 |
3 |
Coding agents are (1) a big deal, (2) very relevant to data leverage, and (3) able to help build tools that support data leverage! 2 uses |
| agent based model |
3-grams |
4 |
2 |
Building a Data Pipeworks for Democratic AI: From Human Knowledge to Records to AI Systems 3 uses |
| ai generated content |
3-grams |
4 |
3 |
Tipping Points for Content Ecosystems 2 uses |
| arxiv preprint arxiv |
3-grams |
4 |
2 |
Public AI, Data Appraisal, and Data Debates 3 uses |
| chatgpt blog post |
3-grams |
4 |
1 |
ChatGPT is Awesome and Scary: You Deserve Credit for the Good Parts (and Might Help Fix the Bad Parts) 4 uses |
| commons rent tax |
3-grams |
4 |
1 |
AI Dividends Without Taxing Compute, Automation, or Equity: A Presumptive Commons-Rent Tax Based on Capabilities and Data Dependence 4 uses |
| data ablation experiments |
3-grams |
4 |
3 |
Public AI, Data Appraisal, and Data Debates 2 uses |
| data leverage publication |
3-grams |
4 |
1 |
How I am publishing right now 4 uses |
| data pipeworks model |
3-grams |
4 |
2 |
Building a Data Pipeworks for Democratic AI: From Human Knowledge to Records to AI Systems 3 uses |
| data protection action |
3-grams |
4 |
2 |
AI Labs Should Open Source Data Protection Technologies 3 uses |
| data relations board |
3-grams |
4 |
2 |
Bing Rewards for the AI Age 3 uses |
| eu ai code |
3-grams |
4 |
1 |
AI Dividends Without Taxing Compute, Automation, or Equity: A Presumptive Commons-Rent Tax Based on Capabilities and Data Dependence 4 uses |
| frontier ai models |
3-grams |
4 |
4 |
A Short Guide to Data Strikes and Conscious Data Contribution in the Context of 2026 Frontier AI 1 use |
| generative ai training |
3-grams |
4 |
2 |
A Harbinger of the Future of Content? The New York Times Starts a Data Strike 2 uses |
| help improve ai |
3-grams |
4 |
2 |
The Coding Agent Data Deal 3 uses |
| human knowledge work |
3-grams |
4 |
2 |
Evaluation Data Leverage: Advances like "Deep Research" Highlight a Looming Opportunity for Bargaining Power 2 uses |
| impact ai capabilities |
3-grams |
4 |
3 |
Plural AI Data Alignment 2 uses |
| instructgpt model card |
3-grams |
4 |
1 |
ChatGPT is Awesome and Scary: You Deserve Credit for the Good Parts (and Might Help Fix the Bad Parts) 4 uses |
| llm social simulation |
3-grams |
4 |
2 |
Two natural allies of a "Data Transparency" agenda: capabilities forecasters and social simulators 3 uses |
| look lot like |
3-grams |
4 |
3 |
The AI "Evaluation Crisis" Is an Opportunity to Get Data Flow Right 2 uses |
| looks lot like |
3-grams |
4 |
4 |
Building a Data Pipeworks for Democratic AI: From Human Knowledge to Records to AI Systems 1 use |
| main data leverage |
3-grams |
4 |
1 |
How I am publishing right now 4 uses |
| model transparency index |
3-grams |
4 |
2 |
AI Dividends Without Taxing Compute, Automation, or Equity: A Presumptive Commons-Rent Tax Based on Capabilities and Data Dependence 3 uses |
| new data rules |
3-grams |
4 |
1 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 4 uses |
| one related attestation |
3-grams |
4 |
1 |
Attestation across the AI Supply Chain 4 uses |
| open knowledge advocate |
3-grams |
4 |
1 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 4 uses |
| post training data |
3-grams |
4 |
4 |
AI Labs Should Open Source Data Protection Technologies 1 use |
| self directed knowledge |
3-grams |
4 |
1 |
Tipping Points for Content Ecosystems 4 uses |
| self supervised pretraining |
3-grams |
4 |
1 |
The AI "Evaluation Crisis" Is an Opportunity to Get Data Flow Right 4 uses |
| stricter data rules |
3-grams |
4 |
1 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 4 uses |
| subsidized club goods |
3-grams |
4 |
1 |
The Paradox of Reuse in 2026: A Case of Quasi-Enclosure, or "Subsidized Club Goods that Sort of Look Like Public Goods" 4 uses |
| substack concrete prediction |
3-grams |
4 |
1 |
How collective bargaining for information, public AI, and HCI research all fit together 4 uses |
| win win win |
3-grams |
4 |
2 |
Attestation across the AI Supply Chain 3 uses |
| academic peer review |
3-grams |
3 |
2 |
AI is driving the cost of polish down; some musings on fancy versus terse artifacts 2 uses |
| agi like ag1 |
3-grams |
3 |
3 |
Selling AGI like AG1: Will Consumers Push Back Against Proprietary Blends of Herbs and of Data? 1 use |
| ai jobs apocalypse |
3-grams |
3 |
1 |
On AI-driven Job Apocalypses and Collective Bargaining for Information 3 uses |
| ai replaces workers |
3-grams |
3 |
1 |
On AI-driven Job Apocalypses and Collective Bargaining for Information 3 uses |
| ai social simulations |
3-grams |
3 |
2 |
How do we know our AI output is good? Double checks, bar charts, vibes, and training data. 2 uses |
| ai systems like |
3-grams |
3 |
3 |
Don’t give OpenAI all the credit for GPT-3: You might have helped create the latest “astonishing” advance in AI too 1 use |
| build ai systems |
3-grams |
3 |
2 |
Don’t give OpenAI all the credit for GPT-3: You might have helped create the latest “astonishing” advance in AI too 2 uses |
| building public ai |
3-grams |
3 |
1 |
Public AI, Data Appraisal, and Data Debates 3 uses |
| choose ai products |
3-grams |
3 |
2 |
"People First" Policy Ideas that Complement Each Other (through better data flow) 2 uses |
| clearer data rules |
3-grams |
3 |
2 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 2 uses |
| coding agent data |
3-grams |
3 |
1 |
The Coding Agent Data Deal 3 uses |
| coding agent paradigm |
3-grams |
3 |
2 |
The Coding Agent Data Deal 2 uses |
| coding agent transcripts |
3-grams |
3 |
2 |
The Coding Agent Data Deal 2 uses |
| consumer decision making |
3-grams |
3 |
2 |
Attestation across the AI Supply Chain 2 uses |
| data contributions flow |
3-grams |
3 |
3 |
Plural AI Data Alignment 1 use |
| data dependence tax |
3-grams |
3 |
1 |
AI Dividends Without Taxing Compute, Automation, or Equity: A Presumptive Commons-Rent Tax Based on Capabilities and Data Dependence 3 uses |
| data dependent computing |
3-grams |
3 |
3 |
ChatGPT is Awesome and Scary: You Deserve Credit for the Good Parts (and Might Help Fix the Bad Parts) 1 use |
| data dependent technologies |
3-grams |
3 |
2 |
A Short Guide to Data Strikes and Conscious Data Contribution in the Context of 2026 Frontier AI 2 uses |
| data leverage power |
3-grams |
3 |
1 |
Will the New York Times Data Strike Have a Large Impact on ChatGPT? 3 uses |
| data mixing work |
3-grams |
3 |
1 |
AI Dividends Without Taxing Compute, Automation, or Equity: A Presumptive Commons-Rent Tax Based on Capabilities and Data Dependence 3 uses |
| data napkin math |
3-grams |
3 |
3 |
Tipping Points for Content Ecosystems 1 use |
| data protection technologies |
3-grams |
3 |
2 |
AI Labs Should Open Source Data Protection Technologies 2 uses |
| data related collective |
3-grams |
3 |
3 |
ChatGPT is Awesome and Scary: You Deserve Credit for the Good Parts (and Might Help Fix the Bad Parts) 1 use |
| data sharing consortia |
3-grams |
3 |
1 |
Public AI, Data Appraisal, and Data Debates 3 uses |
| data strike aimed |
3-grams |
3 |
2 |
Will the New York Times Data Strike Have a Large Impact on ChatGPT? 2 uses |
| data value estimation |
3-grams |
3 |
2 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 2 uses |
| especially good document |
3-grams |
3 |
1 |
Tipping Points for Content Ecosystems 3 uses |
| five stage model |
3-grams |
3 |
1 |
Building a Data Pipeworks for Democratic AI: From Human Knowledge to Records to AI Systems 3 uses |
| global wealth fund |
3-grams |
3 |
1 |
AI Dividends Without Taxing Compute, Automation, or Equity: A Presumptive Commons-Rent Tax Based on Capabilities and Data Dependence 3 uses |
| high quality information |
3-grams |
3 |
2 |
Plural AI Data Alignment 2 uses |
| high quality reference |
3-grams |
3 |
2 |
ChatGPT is Awesome and Scary: You Deserve Credit for the Good Parts (and Might Help Fix the Bad Parts) 2 uses |
| human generated data |
3-grams |
3 |
2 |
Building a Data Pipeworks for Democratic AI: From Human Knowledge to Records to AI Systems 2 uses |
| improve ai systems |
3-grams |
3 |
1 |
The Coding Agent Data Deal 3 uses |
| instructgpt blog post |
3-grams |
3 |
1 |
ChatGPT is Awesome and Scary: You Deserve Credit for the Good Parts (and Might Help Fix the Bad Parts) 3 uses |
| large scale data |
3-grams |
3 |
2 |
Building a Data Pipeworks for Democratic AI: From Human Knowledge to Records to AI Systems 2 uses |
| like stack exchange |
3-grams |
3 |
2 |
Tipping Points for Content Ecosystems 2 uses |
| like wikimedia enterprise |
3-grams |
3 |
2 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 2 uses |
| making ai systems |
3-grams |
3 |
3 |
A Harbinger of the Future of Content? The New York Times Starts a Data Strike 1 use |
| nber working paper |
3-grams |
3 |
2 |
AI Dividends Without Taxing Compute, Automation, or Equity: A Presumptive Commons-Rent Tax Based on Capabilities and Data Dependence 2 uses |
| note taking software |
3-grams |
3 |
1 |
Attestation across the AI Supply Chain 3 uses |
| open source projects |
3-grams |
3 |
2 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 2 uses |
| openai anthropic google |
3-grams |
3 |
2 |
Live by the free-content-for-training sword, die by the free-content-for-training sword 2 uses |
| platforms like stackexchange |
3-grams |
3 |
3 |
The Paradox of Reuse, Language Models Edition 1 use |
| post ai data |
3-grams |
3 |
2 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 2 uses |
| public ai data |
3-grams |
3 |
2 |
How collective bargaining for information, public AI, and HCI research all fit together 2 uses |
| public ai lab |
3-grams |
3 |
2 |
Public AI, Data Appraisal, and Data Debates 2 uses |
| public benefit corporations |
3-grams |
3 |
2 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 2 uses |
| quasi public goods |
3-grams |
3 |
2 |
AI progress as quasi-public good production 2 uses |
| related collective action |
3-grams |
3 |
3 |
ChatGPT is Awesome and Scary: You Deserve Credit for the Good Parts (and Might Help Fix the Bad Parts) 1 use |
| robots txt file |
3-grams |
3 |
2 |
A Harbinger of the Future of Content? The New York Times Starts a Data Strike 2 uses |
| search engine results |
3-grams |
3 |
3 |
Don’t give OpenAI all the credit for GPT-3: You might have helped create the latest “astonishing” advance in AI too 1 use |
| selling agi like |
3-grams |
3 |
3 |
Selling AGI like AG1: Will Consumers Push Back Against Proprietary Blends of Herbs and of Data? 1 use |
| systems like chatgpt |
3-grams |
3 |
3 |
The Paradox of Reuse, Language Models Edition 1 use |
| systems like gpt |
3-grams |
3 |
2 |
Don’t give OpenAI all the credit for GPT-3: You might have helped create the latest “astonishing” advance in AI too 2 uses |
| train ai models |
3-grams |
3 |
3 |
AI Labs Should Open Source Data Protection Technologies 1 use |
| training data like |
3-grams |
3 |
3 |
Bing Rewards for the AI Age 1 use |
| training data transparency |
3-grams |
3 |
3 |
Two natural allies of a "Data Transparency" agenda: capabilities forecasters and social simulators 1 use |
| use coding agent |
3-grams |
3 |
2 |
The Coding Agent Data Deal 2 uses |
| user generated content |
3-grams |
3 |
2 |
Reddit, StackOverflow, and Europe: All Trending Towards Data Dignity 2 uses |
| valuable data labor |
3-grams |
3 |
3 |
Don’t give OpenAI all the credit for GPT-3: You might have helped create the latest “astonishing” advance in AI too 1 use |
| white collar workers |
3-grams |
3 |
2 |
Evaluation Data Leverage: Advances like "Deep Research" Highlight a Looming Opportunity for Bargaining Power 2 uses |
| ablations scaling experiments |
3-grams |
2 |
1 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 2 uses |
| accepted explained data |
3-grams |
2 |
1 |
AI Dividends Without Taxing Compute, Automation, or Equity: A Presumptive Commons-Rent Tax Based on Capabilities and Data Dependence 2 uses |
| across training evaluation |
3-grams |
2 |
1 |
Attestation across the AI Supply Chain 2 uses |
| affects model performance |
3-grams |
2 |
2 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 1 use |
| agent based modeling |
3-grams |
2 |
1 |
Data Leverage Recap: December 2022 - April 2023 2 uses |
| ai art thief |
3-grams |
2 |
2 |
AI Artist or AI Art Thief? Innovation, Public Mandates, and the Case for Talking in Terms of Leverage 1 use |
| ai auditing ecosystem |
3-grams |
2 |
1 |
AI Dividends Without Taxing Compute, Automation, or Equity: A Presumptive Commons-Rent Tax Based on Capabilities and Data Dependence 2 uses |
| ai data alignment |
3-grams |
2 |
2 |
Plural AI Data Alignment 1 use |
| ai data flow |
3-grams |
2 |
2 |
Perplexity CEO's Interaction with Striking New York Times Workers Does Not Reflect Well on the AI Industry 1 use |
| ai data rules |
3-grams |
2 |
1 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 2 uses |
| ai dividend proposal |
3-grams |
2 |
1 |
AI Dividends Without Taxing Compute, Automation, or Equity: A Presumptive Commons-Rent Tax Based on Capabilities and Data Dependence 2 uses |
| ai economic impacts |
3-grams |
2 |
2 |
Building a Data Pipeworks for Democratic AI: From Human Knowledge to Records to AI Systems 1 use |
| ai generated code |
3-grams |
2 |
2 |
"People First" Policy Ideas that Complement Each Other (through better data flow) 1 use |
| ai generated outputs |
3-grams |
2 |
1 |
Attestation across the AI Supply Chain 2 uses |
| ai gets better |
3-grams |
2 |
2 |
On AI-driven Job Apocalypses and Collective Bargaining for Information 1 use |
| ai labs need |
3-grams |
2 |
2 |
Live by the free-content-for-training sword, die by the free-content-for-training sword 1 use |
| ai literacy efforts |
3-grams |
2 |
2 |
Selling AGI like AG1: Will Consumers Push Back Against Proprietary Blends of Herbs and of Data? 1 use |
| ai network website |
3-grams |
2 |
2 |
Public AI, Data Appraisal, and Data Debates 1 use |
| ai operating firms |
3-grams |
2 |
2 |
Bing Rewards for the AI Age 1 use |
| ai public ownership |
3-grams |
2 |
1 |
AI Dividends Without Taxing Compute, Automation, or Equity: A Presumptive Commons-Rent Tax Based on Capabilities and Data Dependence 2 uses |
| ai related policies |
3-grams |
2 |
2 |
Will the New York Times Data Strike Have a Large Impact on ChatGPT? 1 use |
| ai related taxes |
3-grams |
2 |
1 |
AI Dividends Without Taxing Compute, Automation, or Equity: A Presumptive Commons-Rent Tax Based on Capabilities and Data Dependence 2 uses |
| ai safety research |
3-grams |
2 |
2 |
Plural AI Data Alignment 1 use |
| ai social simulation |
3-grams |
2 |
1 |
Two natural allies of a "Data Transparency" agenda: capabilities forecasters and social simulators 2 uses |
| ai systems setting |
3-grams |
2 |
1 |
The Coding Agent Data Deal 2 uses |
| ai training data |
3-grams |
2 |
2 |
A Harbinger of the Future of Content? The New York Times Starts a Data Strike 1 use |
| ai trust stack |
3-grams |
2 |
2 |
"People First" Policy Ideas that Complement Each Other (through better data flow) 1 use |
| ai utility stems |
3-grams |
2 |
2 |
Each Instance of "AI Utility" Stems from Some Human Act(s) of Information Recording and Ranking 1 use |
| allow data use |
3-grams |
2 |
1 |
The Coding Agent Data Deal 2 uses |
| ap ai public |
3-grams |
2 |
1 |
AI Dividends Without Taxing Compute, Automation, or Equity: A Presumptive Commons-Rent Tax Based on Capabilities and Data Dependence 2 uses |
| api keys etc |
3-grams |
2 |
1 |
Coding agents are (1) a big deal, (2) very relevant to data leverage, and (3) able to help build tools that support data leverage! 2 uses |
| appropriate group size |
3-grams |
2 |
1 |
[microblog] One book is worth "0.06%" benchmark points to AI; is "no different from noise". What gives? 2 uses |
| archetypal model builder |
3-grams |
2 |
1 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 2 uses |
| arrow information paradox |
3-grams |
2 |
1 |
Public AI, Data Appraisal, and Data Debates 2 uses |
| art thief innovation |
3-grams |
2 |
2 |
AI Artist or AI Art Thief? Innovation, Public Mandates, and the Case for Talking in Terms of Leverage 1 use |
| artifact without improving |
3-grams |
2 |
1 |
AI is driving the cost of polish down; some musings on fancy versus terse artifacts 2 uses |
| artifacts look nice |
3-grams |
2 |
1 |
AI is driving the cost of polish down; some musings on fancy versus terse artifacts 2 uses |
| artificial general intelligence |
3-grams |
2 |
2 |
Plural AI Data Alignment 1 use |
| artists coders writers |
3-grams |
2 |
2 |
AI Artist or AI Art Thief? Innovation, Public Mandates, and the Case for Talking in Terms of Leverage 1 use |
| attestation forward data |
3-grams |
2 |
2 |
Attestation across the AI Supply Chain 1 use |
| attestation forward strategy |
3-grams |
2 |
1 |
Attestation across the AI Supply Chain 2 uses |
| auditing ecosystem frontier |
3-grams |
2 |
1 |
AI Dividends Without Taxing Compute, Automation, or Equity: A Presumptive Commons-Rent Tax Based on Capabilities and Data Dependence 2 uses |
| augment human intelligence |
3-grams |
2 |
1 |
On AI-driven Job Apocalypses and Collective Bargaining for Information 2 uses |
| author model builder |
3-grams |
2 |
1 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 2 uses |
| bar charts vibes |
3-grams |
2 |
1 |
How do we know our AI output is good? Double checks, bar charts, vibes, and training data. 2 uses |
| bartz anthropic ruling |
3-grams |
2 |
1 |
AI Dividends Without Taxing Compute, Automation, or Equity: A Presumptive Commons-Rent Tax Based on Capabilities and Data Dependence 2 uses |
| based ai labs |
3-grams |
2 |
2 |
Live by the free-content-for-training sword, die by the free-content-for-training sword 1 use |
| based book corpora |
3-grams |
2 |
1 |
ChatGPT is Awesome and Scary: You Deserve Credit for the Good Parts (and Might Help Fix the Bad Parts) 2 uses |
| beige tech microsite |
3-grams |
2 |
1 |
On AI-driven Job Apocalypses and Collective Bargaining for Information 2 uses |
| benefit private actors |
3-grams |
2 |
1 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 2 uses |
| best case scenario |
3-grams |
2 |
2 |
Bing Rewards for the AI Age 1 use |
| better data rules |
3-grams |
2 |
1 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 2 uses |
| blog post details |
3-grams |
2 |
1 |
ChatGPT is Awesome and Scary: You Deserve Credit for the Good Parts (and Might Help Fix the Bad Parts) 2 uses |
| bought reddit data |
3-grams |
2 |
1 |
Public AI, Data Appraisal, and Data Debates 2 uses |
| build powerful ai |
3-grams |
2 |
1 |
Tipping Points for Content Ecosystems 2 uses |
| bullet point summary |
3-grams |
2 |
1 |
Attestation across the AI Supply Chain 2 uses |
| capability relevant data |
3-grams |
2 |
1 |
AI Dividends Without Taxing Compute, Automation, or Equity: A Presumptive Commons-Rent Tax Based on Capabilities and Data Dependence 2 uses |
| capable ai systems |
3-grams |
2 |
2 |
Data Leverage Recap: December 2022 - April 2023 1 use |
| carefully designed incentive |
3-grams |
2 |
1 |
Building a Data Pipeworks for Democratic AI: From Human Knowledge to Records to AI Systems 2 uses |
| cases open data |
3-grams |
2 |
1 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 2 uses |
| cato trump sanders |
3-grams |
2 |
1 |
AI Dividends Without Taxing Compute, Automation, or Equity: A Presumptive Commons-Rent Tax Based on Capabilities and Data Dependence 2 uses |
| check important info |
3-grams |
2 |
1 |
How do we know our AI output is good? Double checks, bar charts, vibes, and training data. 2 uses |
| checks bar charts |
3-grams |
2 |
1 |
How do we know our AI output is good? Double checks, bar charts, vibes, and training data. 2 uses |
| clearly anti social |
3-grams |
2 |
1 |
AI Labs Should Open Source Data Protection Technologies 2 uses |
| co inference utility |
3-grams |
2 |
2 |
How collective bargaining for information, public AI, and HCI research all fit together 1 use |
| coding agent tools |
3-grams |
2 |
1 |
Coding agents are (1) a big deal, (2) very relevant to data leverage, and (3) able to help build tools that support data leverage! 2 uses |
| coding agents create |
3-grams |
2 |
2 |
Coding agents are (1) a big deal, (2) very relevant to data leverage, and (3) able to help build tools that support data leverage! 1 use |
| coding transcript data |
3-grams |
2 |
1 |
Coding agents are (1) a big deal, (2) very relevant to data leverage, and (3) able to help build tools that support data leverage! 2 uses |
| collective bargaining around |
3-grams |
2 |
2 |
Is Zuckerberg right to say that your specific creative work has no value to AI? 1 use |
| community led evaluation |
3-grams |
2 |
1 |
The AI "Evaluation Crisis" Is an Opportunity to Get Data Flow Right 2 uses |
| computational social science |
3-grams |
2 |
2 |
Will the New York Times Data Strike Have a Large Impact on ChatGPT? 1 use |
| consumers push back |
3-grams |
2 |
2 |
Selling AGI like AG1: Will Consumers Push Back Against Proprietary Blends of Herbs and of Data? 1 use |
| content format length |
3-grams |
2 |
1 |
Google and TikTok rank bundles of information; ChatGPT ranks grains. 2 uses |
| contribution evaluation record |
3-grams |
2 |
1 |
Attestation across the AI Supply Chain 2 uses |
| create feedback loops |
3-grams |
2 |
2 |
Building a Data Pipeworks for Democratic AI: From Human Knowledge to Records to AI Systems 1 use |
| cs professor website |
3-grams |
2 |
1 |
AI is driving the cost of polish down; some musings on fancy versus terse artifacts 2 uses |
| current data paradigm |
3-grams |
2 |
2 |
AI Labs Should Open Source Data Protection Technologies 1 use |
| current status quo |
3-grams |
2 |
2 |
Live by the free-content-for-training sword, die by the free-content-for-training sword 1 use |
| data attribution task |
3-grams |
2 |
2 |
The Paradox of Reuse, Language Models Edition 1 use |
| data centric ai |
3-grams |
2 |
2 |
Building a Data Pipeworks for Democratic AI: From Human Knowledge to Records to AI Systems 1 use |
| data centric research |
3-grams |
2 |
2 |
Two natural allies of a "Data Transparency" agenda: capabilities forecasters and social simulators 1 use |
| data deals tracker |
3-grams |
2 |
2 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 1 use |
| data dividend report |
3-grams |
2 |
1 |
AI Dividends Without Taxing Compute, Automation, or Equity: A Presumptive Commons-Rent Tax Based on Capabilities and Data Dependence 2 uses |
| data dividends report |
3-grams |
2 |
1 |
AI Dividends Without Taxing Compute, Automation, or Equity: A Presumptive Commons-Rent Tax Based on Capabilities and Data Dependence 2 uses |
| data flow building |
3-grams |
2 |
2 |
Building a Data Pipeworks for Democratic AI: From Human Knowledge to Records to AI Systems 1 use |
| data flow problem |
3-grams |
2 |
2 |
The AI "Evaluation Crisis" Is an Opportunity to Get Data Flow Right 1 use |
| data generating activity |
3-grams |
2 |
1 |
Evaluation Data Leverage: Advances like "Deep Research" Highlight a Looming Opportunity for Bargaining Power 2 uses |
| data hungry ai |
3-grams |
2 |
2 |
Reddit, StackOverflow, and Europe: All Trending Towards Data Dignity 1 use |
| data intermediary organizations |
3-grams |
2 |
2 |
AI Labs Should Open Source Data Protection Technologies 1 use |
| data leverage angle |
3-grams |
2 |
1 |
AI is driving the cost of polish down; some musings on fancy versus terse artifacts 2 uses |
| data leverage campaigns |
3-grams |
2 |
1 |
A Short Guide to Data Strikes and Conscious Data Contribution in the Context of 2026 Frontier AI 2 uses |
| data leverage posts |
3-grams |
2 |
2 |
Attestation across the AI Supply Chain 1 use |
| data leverage right |
3-grams |
2 |
1 |
Evaluation Data Leverage: Advances like "Deep Research" Highlight a Looming Opportunity for Bargaining Power 2 uses |
| data licensing contract |
3-grams |
2 |
2 |
Public AI, Data Appraisal, and Data Debates 1 use |
| data mixture experiments |
3-grams |
2 |
1 |
AI Dividends Without Taxing Compute, Automation, or Equity: A Presumptive Commons-Rent Tax Based on Capabilities and Data Dependence 2 uses |
| data protection techniques |
3-grams |
2 |
1 |
AI Labs Should Open Source Data Protection Technologies 2 uses |
| data related social |
3-grams |
2 |
1 |
Building a Data Pipeworks for Democratic AI: From Human Knowledge to Records to AI Systems 2 uses |
| data rules clear |
3-grams |
2 |
1 |
AI Dividends Without Taxing Compute, Automation, or Equity: A Presumptive Commons-Rent Tax Based on Capabilities and Data Dependence 2 uses |
| data rules piece |
3-grams |
2 |
1 |
Attestation across the AI Supply Chain 2 uses |
| data strike impact |
3-grams |
2 |
1 |
Will the New York Times Data Strike Have a Large Impact on ChatGPT? 2 uses |
| data transparency laws |
3-grams |
2 |
2 |
Will the New York Times Data Strike Have a Large Impact on ChatGPT? 1 use |
| data valuation experiments |
3-grams |
2 |
2 |
Tipping Points for Content Ecosystems 1 use |
| data valuation infrastructure |
3-grams |
2 |
1 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 2 uses |
| data value estimates |
3-grams |
2 |
1 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 2 uses |
| data workers inquiry |
3-grams |
2 |
2 |
Attestation across the AI Supply Chain 1 use |
| de facto carve |
3-grams |
2 |
1 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 2 uses |
| decision making around |
3-grams |
2 |
2 |
Public AI, Data Appraisal, and Data Debates 1 use |
| deep research agents |
3-grams |
2 |
1 |
Evaluation Data Leverage: Advances like "Deep Research" Highlight a Looming Opportunity for Bargaining Power 2 uses |
| deep research release |
3-grams |
2 |
1 |
Evaluation Data Leverage: Advances like "Deep Research" Highlight a Looming Opportunity for Bargaining Power 2 uses |
| deployed ai systems |
3-grams |
2 |
1 |
Building a Data Pipeworks for Democratic AI: From Human Knowledge to Records to AI Systems 2 uses |
| design decisions etc |
3-grams |
2 |
2 |
Public AI, Data Appraisal, and Data Debates 1 use |
| designed incentive systems |
3-grams |
2 |
1 |
Building a Data Pipeworks for Democratic AI: From Human Knowledge to Records to AI Systems 2 uses |
| determining ai capabilities |
3-grams |
2 |
1 |
Building a Data Pipeworks for Democratic AI: From Human Knowledge to Records to AI Systems 2 uses |
| directed knowledge contribution |
3-grams |
2 |
1 |
Tipping Points for Content Ecosystems 2 uses |
| dispossession economic harms |
3-grams |
2 |
1 |
A Harbinger of the Future of Content? The New York Times Starts a Data Strike 2 uses |
| distribute utility widely |
3-grams |
2 |
1 |
The Paradox of Reuse in 2026: A Case of Quasi-Enclosure, or "Subsidized Club Goods that Sort of Look Like Public Goods" 2 uses |
| dividend proposal sanders |
3-grams |
2 |
1 |
AI Dividends Without Taxing Compute, Automation, or Equity: A Presumptive Commons-Rent Tax Based on Capabilities and Data Dependence 2 uses |
| dividends report data |
3-grams |
2 |
1 |
AI Dividends Without Taxing Compute, Automation, or Equity: A Presumptive Commons-Rent Tax Based on Capabilities and Data Dependence 2 uses |
| double check every |
3-grams |
2 |
1 |
How do we know our AI output is good? Double checks, bar charts, vibes, and training data. 2 uses |
| double checks bar |
3-grams |
2 |
1 |
How do we know our AI output is good? Double checks, bar charts, vibes, and training data. 2 uses |
| economic harms via |
3-grams |
2 |
1 |
A Harbinger of the Future of Content? The New York Times Starts a Data Strike 2 uses |
| economically valuable work |
3-grams |
2 |
1 |
On AI-driven Job Apocalypses and Collective Bargaining for Information 2 uses |
| ecosystem frontier ai |
3-grams |
2 |
1 |
AI Dividends Without Taxing Compute, Automation, or Equity: A Presumptive Commons-Rent Tax Based on Capabilities and Data Dependence 2 uses |
| edited english wikipedia |
3-grams |
2 |
2 |
Don’t give OpenAI all the credit for GPT-3: You might have helped create the latest “astonishing” advance in AI too 1 use |
| effective altruism forum |
3-grams |
2 |
1 |
On AI-driven Job Apocalypses and Collective Bargaining for Information 2 uses |
| enable certain types |
3-grams |
2 |
1 |
The Paradox of Reuse in 2026: A Case of Quasi-Enclosure, or "Subsidized Club Goods that Sort of Look Like Public Goods" 2 uses |
| enable collective bargaining |
3-grams |
2 |
2 |
How collective bargaining for information, public AI, and HCI research all fit together 1 use |
| english language wikipedia |
3-grams |
2 |
1 |
ChatGPT is Awesome and Scary: You Deserve Credit for the Good Parts (and Might Help Fix the Bad Parts) 2 uses |
| enterprise wikimedia enterprise |
3-grams |
2 |
1 |
AI Dividends Without Taxing Compute, Automation, or Equity: A Presumptive Commons-Rent Tax Based on Capabilities and Data Dependence 2 uses |
| et al arxiv |
3-grams |
2 |
1 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 2 uses |
| eval data leverage |
3-grams |
2 |
2 |
How do we know our AI output is good? Double checks, bar charts, vibes, and training data. 1 use |
| evaluating deep research |
3-grams |
2 |
1 |
Evaluation Data Leverage: Advances like "Deep Research" Highlight a Looming Opportunity for Bargaining Power 2 uses |
| evaluation record output |
3-grams |
2 |
1 |
Attestation across the AI Supply Chain 2 uses |
| every frontier model |
3-grams |
2 |
2 |
Two natural allies of a "Data Transparency" agenda: capabilities forecasters and social simulators 1 use |
| examples please let |
3-grams |
2 |
2 |
AI Technologies are System Maps, and You are a Cartographer 1 use |
| explained data points |
3-grams |
2 |
1 |
AI Dividends Without Taxing Compute, Automation, or Equity: A Presumptive Commons-Rent Tax Based on Capabilities and Data Dependence 2 uses |
| facto carve outs |
3-grams |
2 |
1 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 2 uses |
| film noir fans |
3-grams |
2 |
1 |
Will the New York Times Data Strike Have a Large Impact on ChatGPT? 2 uses |
| fine grained data |
3-grams |
2 |
2 |
Evaluation Data Leverage: Advances like "Deep Research" Highlight a Looming Opportunity for Bargaining Power 1 use |
| firms like mercor |
3-grams |
2 |
1 |
The AI "Evaluation Crisis" Is an Opportunity to Get Data Flow Right 2 uses |
| first step towards |
3-grams |
2 |
2 |
Plural AI Data Alignment 1 use |
| fit models stage |
3-grams |
2 |
1 |
Building a Data Pipeworks for Democratic AI: From Human Knowledge to Records to AI Systems 2 uses |
| fortune sacks ai |
3-grams |
2 |
1 |
AI Dividends Without Taxing Compute, Automation, or Equity: A Presumptive Commons-Rent Tax Based on Capabilities and Data Dependence 2 uses |
| forward data strategy |
3-grams |
2 |
2 |
Attestation across the AI Supply Chain 1 use |
| free tier users |
3-grams |
2 |
2 |
The Coding Agent Data Deal 1 use |
| full supply chain |
3-grams |
2 |
1 |
Attestation across the AI Supply Chain 2 uses |
| funds nber ai |
3-grams |
2 |
1 |
AI Dividends Without Taxing Compute, Automation, or Equity: A Presumptive Commons-Rent Tax Based on Capabilities and Data Dependence 2 uses |
| future data creation |
3-grams |
2 |
2 |
Tipping Points for Content Ecosystems 1 use |
| generative ai tools |
3-grams |
2 |
1 |
Google and TikTok rank bundles of information; ChatGPT ranks grains. 2 uses |
| get 6b tokens |
3-grams |
2 |
1 |
Tipping Points for Content Ecosystems 2 uses |
| giant ranking exercise |
3-grams |
2 |
1 |
Google and TikTok rank bundles of information; ChatGPT ranks grains. 2 uses |
| giving people agency |
3-grams |
2 |
2 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 1 use |
| going around house |
3-grams |
2 |
1 |
Building a Data Pipeworks for Democratic AI: From Human Knowledge to Records to AI Systems 2 uses |
| group level data |
3-grams |
2 |
2 |
Will the New York Times Data Strike Have a Large Impact on ChatGPT? 1 use |
| growth investment diffusion |
3-grams |
2 |
1 |
AI Dividends Without Taxing Compute, Automation, or Equity: A Presumptive Commons-Rent Tax Based on Capabilities and Data Dependence 2 uses |
| harms via competition |
3-grams |
2 |
1 |
A Harbinger of the Future of Content? The New York Times Starts a Data Strike 2 uses |
| harry potter books |
3-grams |
2 |
1 |
Will the New York Times Data Strike Have a Large Impact on ChatGPT? 2 uses |
| harvard business review |
3-grams |
2 |
1 |
On AI-driven Job Apocalypses and Collective Bargaining for Information 2 uses |
| hayes et al |
3-grams |
2 |
1 |
How do we know our AI output is good? Double checks, bar charts, vibes, and training data. 2 uses |
| help prevent ai |
3-grams |
2 |
1 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 2 uses |
| high quality data |
3-grams |
2 |
2 |
Reddit, StackOverflow, and Europe: All Trending Towards Data Dignity 1 use |
| high value contributions |
3-grams |
2 |
1 |
The Paradox of Reuse in 2026: A Case of Quasi-Enclosure, or "Subsidized Club Goods that Sort of Look Like Public Goods" 2 uses |
| higher stakes domains |
3-grams |
2 |
1 |
Attestation across the AI Supply Chain 2 uses |
| highly capable ai |
3-grams |
2 |
2 |
A Short Guide to Data Strikes and Conscious Data Contribution in the Context of 2026 Frontier AI 1 use |
| idealized supervised learning |
3-grams |
2 |
1 |
The AI "Evaluation Crisis" Is an Opportunity to Get Data Flow Right 2 uses |
| implement something like |
3-grams |
2 |
2 |
Data Leverage Recap: December 2022 - April 2023 1 use |
| individual data creators |
3-grams |
2 |
2 |
Reddit, StackOverflow, and Europe: All Trending Towards Data Dignity 1 use |
| industry self governance |
3-grams |
2 |
1 |
Attestation across the AI Supply Chain 2 uses |
| information retrieval systems |
3-grams |
2 |
1 |
Bing Rewards for the AI Age 2 uses |
| informational playing field |
3-grams |
2 |
1 |
Public AI, Data Appraisal, and Data Debates 2 uses |
| innovation public mandates |
3-grams |
2 |
2 |
AI Artist or AI Art Thief? Innovation, Public Mandates, and the Case for Talking in Terms of Leverage 1 use |
| internet based book |
3-grams |
2 |
1 |
ChatGPT is Awesome and Scary: You Deserve Credit for the Good Parts (and Might Help Fix the Bad Parts) 2 uses |
| ip issue model |
3-grams |
2 |
1 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 2 uses |
| issue model builders |
3-grams |
2 |
1 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 2 uses |
| jacob thebault spieker |
3-grams |
2 |
2 |
Tipping Points for Content Ecosystems 1 use |
| kaplan scaling laws |
3-grams |
2 |
1 |
AI Dividends Without Taxing Compute, Automation, or Equity: A Presumptive Commons-Rent Tax Based on Capabilities and Data Dependence 2 uses |
| key data source |
3-grams |
2 |
1 |
ChatGPT is Awesome and Scary: You Deserve Credit for the Good Parts (and Might Help Fix the Bad Parts) 2 uses |
| kittur et al |
3-grams |
2 |
2 |
Bing Rewards for the AI Age 1 use |
| knowledge free culture |
3-grams |
2 |
1 |
Almost Everybody -- Including Both Data Creators and AI Companies -- Stands to Benefit from Clearer "Data Rules". 2 uses |