iiinova — where your digital self keeps evolving

Ruo Shui @ruoshuiresearch.bsky.social · 15h

"It’s a form of performative staging that consists of saying: 'Save us, regulation needs to slow things down because we don’t know what we’re doing.' And at the same time, they would like to have total control over how they are regulated."

060

Ruo Shui @ruoshuiresearch.bsky.social · 15h

"Use is more common among students from non-English-speaking countries and at lower-ranked programs."

041

Reposted by Ruo Shui

Adam @adamshap.bsky.social · 10/10/2026

30% of stem phd dissertation abstracts written by AI

195

Reposted by Ruo Shui

Nathan Lambert @natolambert.bsky.social · 10/10/2026

At the end of this era I'll be the only open models person Nvidia hasn't acquired

4461

Ruo Shui @ruoshuiresearch.bsky.social · 16h

Something that we are learning is that the concept of FU money is wrong. If anything, it’s been the world’s richest people that have debased themselves more than anyone

3237

Reposted by Ruo Shui

SE Gyges @segyges.bsky.social · 11/10/2026

open source and open data as an ethos is largely anti capital and anti authority and is inherently political in ways we need more of

1155977

Ruo Shui @ruoshuiresearch.bsky.social · 16h

people not interested in interpretability are usually uninteresting people

140

Ruo Shui @ruoshuiresearch.bsky.social · 09/10/2026

really cool to make code and data available like this so anyone can try to replicate and comb through the analysis themselves

130

Reposted by Ruo Shui

Ben Glasner @benglasner.bsky.social · 09/10/2026

It remains shocking to me that we have seen a full scale reversal of the post-COVID trend in wage growth. Since March 2025, real wages are down across the distribution. Only the 25th is up (+1%). let’s look at how odd this is in the context of the last 40+ years. @innovateeconomy.bsky.social

1133

Reposted by Ruo Shui

amos @fasterthanli.me · 09/10/2026

Honeymoon with Astra is over btw :/ if they truly don’t fuck with quantized variants based on time of day / release schedule then woo, humans are bad at estimating model quality huh. Maybe I should stick with Chinese models for ML work?

5351

Reposted by Ruo Shui

Conor Sen @conorsen.bsky.social · 09/10/2026

New from me: The labor market has strengthened but like the stock market it’s become narrower and more dependent on the AI buildout conorsen.substack.com/p/the-labor-...

1336

Reposted by Ruo Shui

Daniel Zhao @danielzhao.bsky.social · 13/08/2026

More data from Indeed on data center jobs: Pay premiums are pretty substantial for data center jobs vs. other jobs with a similar job title. Additionally, benefits look to be more generous too. Source: hiringlab.indeed.com/2026/08/13/w...

Faceted dumbbell chart titled “Data center roles pay a premium over the same job elsewhere” shows median posted pay for data center vs. non-data center postings with the same job title taken from January to June 2026. The figure is divided into an upper panel detailing annual pay and a lower panel detailing hourly pay. The largest premium is for a facilities manager (63%).

Heatmap-styled chart titled “Data center jobs advertise more benefits” shows how much more often a US data center posting advertises each benefit (Health, Dental, 401k, PTO, Parental Leave, Relocation Assistance, and a Signing Bonus) compared to a non-data center posting in each occupation, measured over January to June 2026, where 1x = the same rate of advertising.

031

Reposted by Ruo Shui

Axios @axios.com · 09/10/2026

SCOOP: Top executives at Anthropic, OpenAI and other AI companies are privately gaming out scenarios for a public and political revolt after a catastrophic AI event.

axios.com

Scoop: AI companies plot "day after" scenarios for public revolt

Inside the AI industry's worst-case-scenario prep.

125029

Ruo Shui @ruoshuiresearch.bsky.social · 09/10/2026

They wrote up a letter talking about how they believe they were fired for prioritizing safety over the near-term interests of OpenAI as a corporation

OpenAI cannot make AI safe on its own

A note on third-party collaborations, open debate, and clear operating procedures

Dear Safety and Security Committee, Safety Advisory Group, and Mission Advisory Council:

We are the three safety and alignment employees who were fired from OpenAI last week: Tomek Korbak, Jasmine Wang, and Mikita Balesni. We are writing to you because you hold oversight responsibility for safety at OpenAI; our firing and the way it was handled bear directly on it.

We have become concerned that internal and external communications around our firing have made our former colleagues afraid to speak and operate in ways that, until last week, were an integral part of working at OpenAI. We could raise safety concerns and disagree openly, and were encouraged to draw on the expertise of independent safety organizations. This is part of what made OpenAI special, and why we are immensely proud to have been part of the team.

AI is not a normal technology, and OpenAI is not a normal company. Those of us who work on safety see risks before anyone else, and we rely on close collaboration with outside experts to work out how to address them. The freedom to do so without fear, and to have well-defined internal procedures that enable this work, is itself an essential safety mechanism.

We do not believe the path to superintelligence can be navigated safely if the people closest to the risks can no longer work in high-trust, high-bandwidth ways with each other and with third parties. It is that culture we are writing to defend.

1285

Reposted by Ruo Shui

Eugene Vinitsky 🍒 @eugenevinitsky.bsky.social · 08/10/2026

it only now hit me that I'm going to have to write "super intelligence" in all my grants now

151806

Reposted by Ruo Shui

Nick Kodama 🌳👻 @nickkodama.bsky.social · 08/10/2026

LLM (Large Line Man) solves math problem

212373474

Ruo Shui @ruoshuiresearch.bsky.social · 08/10/2026

"productivity" not found in this speech

000

Ruo Shui @ruoshuiresearch.bsky.social · 08/10/2026

this headline won't even last the week

180

Ruo Shui @ruoshuiresearch.bsky.social · 08/10/2026

Furman joins the crowd out takes

140

Ruo Shui @ruoshuiresearch.bsky.social · 08/10/2026

bond bulls gonna be talking Economist cover contra

030

Ruo Shui @ruoshuiresearch.bsky.social · 08/10/2026

"In one Claude Code session, the user asked Claude to create a security researcher résumé that specifically included bullet points conveying the results of the ARTEX-related activity. The prompt included the following personal details..."

crowdstrike.com

Unknown Threat Actor Uses AI-Driven ARTEX to Target South Korean Finance

CrowdStrike Intelligence identified infrastructure associated with a targeted campaign against South Korean financial institutions that resulted in exfiltrated data.

040

Reposted by Ruo Shui

Mark Riedl @markriedl.bsky.social · 06/10/2026

The Capabilibara project seeks to understand how a language model learns to interpret people's beliefs, emotions, intentions, and everyday moral choices. We trace that ability back to the training data using influence functions and unlearning hcai-lab-gt.github.io/capabilibara/ Find us at COLM!

1358

Reposted by Ruo Shui

Catherine Rampell @crampell.bsky.social · 07/10/2026

Trump admin's new rule adding prohibitively expensive new fees for "OPT"—temporary work authorization program for international students and new grads in the United States—is out. Would require colleges to pay $70,000 for year one. $30,000 for years two and three combined for STEM fields.

public-inspection.federalregister.gov

6329152

Ruo Shui @ruoshuiresearch.bsky.social · 07/10/2026

Adobe has been struggling for a while and this won't help

040

Reposted by Ruo Shui

Eugene Vinitsky 🍒 @eugenevinitsky.bsky.social · 07/10/2026

Almost all major media (LLMs included) is now owned by a very small number of people / governments. An exciting propaganda war looms on the horizon

8809

Reposted by Ruo Shui

Patrick De Haan @gasbuddyguy.bsky.social · 07/10/2026

TS Isaias is now strengthening, shifting E. Frcst Cat 2-3 on landfall, impact ~1.5mbpd of refining capacity- primary concerns flooding and winds. Some refiners in the path will likely have to start reducing throughput very soon. Could impact prices- watching- will update.

08018

Ruo Shui @ruoshuiresearch.bsky.social · 07/10/2026

"Investors who SpaceX has previously approached about financing its multibillion-dollar chip purchase said they only received a short two-page deal memo with pictures of outer space and an arrow pointing out that the company was going to build data centres 'somewhere in the universe'."

ft.com

SpaceX looks to raise $40bn to buy Nvidia chips

Blockbuster debt deal is the latest sign of the vast spending on chips and other infrastructure underpinning AI

220

Ruo Shui @ruoshuiresearch.bsky.social · 07/10/2026

Counterfactual line without tariffs (blue dashed line)

A line chart titled "Tariffs Raised Consumer Goods Price Inflation" showing consumer price inflation percentage from January 2022 to early 2026.

Inflation peaked above 9% in mid-2022, steadily dropped below 0% by 2024, and began diverging in early 2025.

Actual Line (Red Solid Line): Rises back above 0% in 2025, reaching around 2% by early 2026.

Counterfactual Line without Tariffs (Blue Dashed Line): Drops below 0% in early 2025 and remains negative through early 2026.

Shaded Area: Highlights the gap between actual inflation and counterfactual inflation without tariffs.

Sources & Notes: Cites data from the U.S. Bureau of Labor Statistics, Census Bureau, USITC, and BEA. Notes clarify that the solid line represents the 12-month change in prices across 67 non-oil goods categories, and the dashed line subtracts estimated tariff contributions.

0145

Ruo Shui @ruoshuiresearch.bsky.social · 06/10/2026

gonna need this same survey for AI

A line graph titled "Some nations are now as negative as the U.S. about social media's impact on democracy," showing the percentage of respondents who say social media has been "more of a bad thing" for democracy in their country between 2022 and 2026.

U.S. (light blue line) remains flat at around 64% from 2022 to 2026. Netherlands (red line) rises significantly from just over 50% in 2022 to nearly 70% in 2026, surpassing all other listed countries.

France (yellow line) increases from roughly 50% to equal the U.S. at around 64%.

Germany (white line) shows a sharp increase from about 40% in 2022 to over 60% in 2026.

U.K. (orange line) and Canada (purple line) both rise from around 48–50% to roughly 54–55%.

Several other unnamed nations (faded grey lines) are also shown increasing from lower percentages across the same period.

170

Reposted by Ruo Shui

vegavandal.bsky.social @vegavandal.bsky.social · 06/10/2026

Must be frustrating this year to be a software investor and have the terminal value of your co in question when like IREN is more likely than not a donut by 2031

1132

Reposted by Ruo Shui

Joe Brusuelas @joebrusuelas.bsky.social · 06/10/2026

US Trade Deficit: take a look at the August increase in the The primary driver is imports of industrial supplies to support the cap ex buildout of AI. Imports were up 4.3% on the month, exports 1.6% after falling over the previous two months. #Econ

0113