TR

Chad Jones ✓

@ChadJonesEcon · joined 27 May 2020

Economics professor at Stanford GSB working on AI and economic growth.

16 976Followers

948Following

253Posts total

1.1MViews on collected posts

Son gönderiler

Chad Jones ✓@ChadJonesEcon10 Aug

Very happy to see "AI and Our Economic Future" published. Thanks to the team at the JEP, especially @TimothyTTaylor and @heidilwilliams_ . https://t.co/Rc8OWL6BlN 25.3K views · 292 likes · 62 reposts · 5 replies Open on X →

Chad Jones ✓@ChadJonesEcon23 Jun

@ChadJonesEcon Dear Chad, please publish your exact agreement with Anthropic regarding your independence as a researcher, NDAs, etc. The public has the right to know this. Thank you. 7.8K views · 48 likes · 0 reposts · 1 replies Open on X →

Chad Jones ✓@ChadJonesEcon23 Jun

Quote: "The Anthropic Institute’s goal is to tell the world what we’re learning about these challenges as we build frontier AI systems, and to partner with external audiences to help address the risks" https://t.co/nZhv9nv9V2 2/3 44.6K views · 265 likes · 19 reposts · 6 replies Open on X →

Chad Jones ✓@ChadJonesEcon23 Jun

Very excited to work with @akorinek, @PeterMcCrory, @jackclarkSF, and the many talented people at @AnthropicAI. 3/3 25.3K views · 233 likes · 9 reposts · 9 replies Open on X →

Chad Jones ✓@ChadJonesEcon23 Jun

Some news: As of June 30, I'll be on leave from Stanford at Anthropic. I'm joining the Anthropic Institute, where I'll continue my research on AI and our economic future and give seminars and talks as always. 1/3 865.6K views · 2.4K likes · 169 reposts · 94 replies Open on X →

Chad Jones ✓@ChadJonesEcon15 Jun

A recent video presentation of "AI and Our Economic Future" from the Stanford GSB Alumni Reunion (i.e. for a broad audience): https://t.co/RqR013Atdn 46.4K views · 177 likes · 58 reposts · 9 replies Open on X →

Chad Jones ✓@ChadJonesEcon16 Apr

Great writeup, Jacob. Thanks! 6.9K views · 30 likes · 4 reposts · 3 replies Open on X →

Chad Jones ✓@ChadJonesEcon16 Apr

Attended @ChadJonesEcon's lecture in Berkeley on AI and economic growth. His "weak links" model shows AI's benefits take decades to compound, but the harms don't wait. That asymmetry bridges the safety vs. acceleration debate in a way neither side's priors alone can generate. h 15.8K views · 87 likes · 14 reposts · 4 replies Open on X →

Chad Jones ✓@ChadJonesEcon12 Jan

@ChadJonesEcon @ChrisTonetti Prof. Jones, it’s much more readable if you put the tweets in order — each tweet replying to the one before it. You can do this so in the app or on browser so that they all send at the same time. This would greatly improve the threat, which is confusi 896 views · 19 likes · 0 reposts · 1 replies Open on X →

Chad Jones ✓@ChadJonesEcon12 Jan

@ChrisTonetti Why? Weak links. We automate more tasks and the rate at which machines get better accelerates — machines are used to improve machines. Nevertheless, the economy is always constrained by the weak links performed by slowly-improving humans. Explosions can be slow! N 1.5K views · 12 likes · 0 reposts · 1 replies Open on X →

Chad Jones ✓@ChadJonesEcon12 Jan

@ChrisTonetti Second, the explosion is surprisingly slow. After 20 years, we are only 4% richer than with constant growth. After 40 years, we are only 19% richer. 17/N https://t.co/aOwZbKM3f7 4.6K views · 24 likes · 1 reposts · 0 replies Open on X →

Chad Jones ✓@ChadJonesEcon12 Jan

@ChrisTonetti TWO SURPRISE TAKEAWAYS: First, for the next 75 years, the macroeconomy looks very similar in all three scenarios, and growth accelerates during those 75 years. Growth in the next half century may be less informative than we thought about long-run outcomes. 16/N 932 views · 9 likes · 0 reposts · 0 replies Open on X →

Chad Jones ✓@ChadJonesEcon12 Jan

@ChrisTonetti If some tasks must always be performed by labor, growth eventually falls to the slow rate at which humans get better (an AL model with a 100% labor share). Surprisingly, growth also explodes in the intermediate case where labor is always used and paid 63% of GDP. 1 887 views · 8 likes · 0 reposts · 0 replies Open on X →

Chad Jones ✓@ChadJonesEcon12 Jan

@ChrisTonetti What happens to growth in the three scenarios? If all tasks get automated, growth explodes with infinite income in finite time (an AK model with a 100% capital share). 14/N https://t.co/a0ujWLFOAJ 1.1K views · 7 likes · 1 reposts · 0 replies Open on X →

Chad Jones ✓@ChadJonesEcon12 Jan

@ChrisTonetti Here's a graph of the share of GDP paid to capital in those three scenarios. Interestingly, the intermediate case shows that the capital share can be constant even though humans perform a vanishing number of tasks. 13/N https://t.co/8jik8lZvIJ 1.1K views · 11 likes · 0 reposts · 0 replies Open on X →

Chad Jones ✓@ChadJonesEcon12 Jan

@ChrisTonetti So we simulate three possibilities: all tasks automated in finite time, some tasks restricted to humans forever, and an intermediate case in which humans perform a share of tasks that vanishes over time. 12/N 938 views · 7 likes · 0 reposts · 0 replies Open on X →

Chad Jones ✓@ChadJonesEcon12 Jan

@ChrisTonetti As is broadly appreciated, whether *all* tasks are ultimately automated versus some tasks *always* restricted to humans matters for very long run outcomes. 11/N 968 views · 9 likes · 1 reposts · 0 replies Open on X →

Chad Jones ✓@ChadJonesEcon12 Jan

@ChrisTonetti The real gains come when automating software helps us to further automate other tasks, including the production of ideas. 10/N 1K views · 12 likes · 0 reposts · 0 replies Open on X →

Chad Jones ✓@ChadJonesEcon12 Jan

@ChrisTonetti If we had infinite output of software on tasks currently performed by software, how much richer would we be? Only about 2 percent richer! The *other* weak links hold down output. 9/N 1.2K views · 12 likes · 2 reposts · 0 replies Open on X →

Chad Jones ✓@ChadJonesEcon12 Jan

@ChrisTonetti Weak links are crucial: Tasks are complements in our model, so even infinite output of some tasks does not lead to infinite GDP. The economy is constrained by the weakest links. 8/N 1.2K views · 10 likes · 1 reposts · 1 replies Open on X →

Chad Jones ✓@ChadJonesEcon12 Jan

@ChrisTonetti FUTURE IMPLICATIONS OF A.I.? We build an idea-driven growth model in which the production of goods and ideas are endogenously automated. We calibrate to our historical evidence, viewing A.I. as the next step on our continuing path of growth through automation. 7/N 1.1K views · 10 likes · 0 reposts · 0 replies Open on X →

Chad Jones ✓@ChadJonesEcon12 Jan

@ChrisTonetti Main takeaway from the first part: Growth results from using rapidly improving machines instead of slowly improving humans on an increasing set of tasks. 6/N 1.2K views · 11 likes · 0 reposts · 0 replies Open on X →

Chad Jones ✓@ChadJonesEcon12 Jan

@ChrisTonetti Even with this rapid capital productivity growth, had we frozen the set of tasks that were automated in 1950, more than half of economic growth would be lost. Because of weak links, machines getting better on a fixed set of tasks does not deliver much growth. 5/N 1.3K views · 12 likes · 1 reposts · 0 replies Open on X →

Chad Jones ✓@ChadJonesEcon12 Jan

@ChrisTonetti Past growth is largely due to endogenous automation that results from machines getting better rapidly. Average annual growth in capital productivity is at least 5pp greater than the average rate at which humans get better at performing tasks. 4/N 1.4K views · 12 likes · 0 reposts · 0 replies Open on X →

Chad Jones ✓@ChadJonesEcon12 Jan

@ChrisTonetti Back to the beginning: First, we perform an accounting exercise to study "How much of past economic growth was due to automation?" Second, we use what we learn to study A.I.'s future economic impact. 3/N 1.7K views · 12 likes · 0 reposts · 0 replies Open on X →

Chad Jones ✓@ChadJonesEcon12 Jan

@ChrisTonetti A preview of where we end up: here's a provocative graph that shows our simulation of the future. Growth explodes, but the explosion is surprisingly slow because of weak links. 2/N https://t.co/v7GDmz34Ul 3.8K views · 39 likes · 3 reposts · 0 replies Open on X →

Chad Jones ✓@ChadJonesEcon12 Jan

New paper (work in progress): "Past Automation and Future AI: How Weak Links Tame the Growth Explosion" with @ChrisTonetti. https://t.co/8f1hQsANNC 1/N 84.2K views · 271 likes · 56 reposts · 22 replies Open on X →

Aynı büyüklükteki hesaplara karşı

Son 90 güne ait 1 gönderi, 10K–100K takipçi aralığıyla yan yana. aynı büyüklükteki hesaplardan daha çok kişiye gösteriliyor.

Medyan görüntülenme25 332bu hesap92410K–100K için medyan

Erişim, %149.22%bu hesap3.62%10K–100K için medyan

Etkileşim, %1.44%bu hesap1.52%10K–100K için medyan

Ölçüt	Bu hesap	10K–100K için medyan	Oran
Gönderi başına medyan görüntülenme	25 332	924	27.4×
Erişim (görüntülenme ÷ takipçi)	149.22%	3.62%	41.2×
Etkileşim oranı	1.44%	1.52%	0.95×

Bu aralıktaki diğer hesaplar → Başka bir hesapla karşılaştır → Bu kıyas değerleri nasıl kuruluyor →

Growth & engagement

How the posts we collected actually performed: views and reaction rate post by post, what the audience did with them, and where the follower count goes.

Views per post

1.1K12 Jan

887

932

4.6K

1.5K

896

15.8K16 Apr

6.9K

46.4K15 Jun

865.6K23 Jun

25.3K

44.6K

7.8K

25.3K10 Aug

Last 14 collected posts, oldest on the left. The scale is logarithmic: one post can outrun the rest a hundred times over.

Engagement rate per post

0.81%12 Jan

0.90%

0.97%

0.54%

0.96%

2.23%

0.68%16 Apr

0.54%

0.53%15 Jun

0.31%23 Jun

0.99%

0.65%

0.63%

1.44%10 Aug

Reactions — likes, reposts, replies and quotes — divided by views. Median for 10K–100K accounts is 1.52%.

What the audience does

Likes68.5%4 024 in total

Reposts6.8%401 in total

Replies2.7%156 in total

Quotes1.6%95 in total

Bookmarks20.4%1 196 in total

Share of every reaction we collected for this account. Replies mean argument, reposts mean endorsement, bookmarks mean the post was worth keeping.

The follower curve appears once this account has two daily snapshots — we take one a day, and this one is on its first.

Benzer hesaplar