Artificial Intelligence: News, Business, Science

Sep 8, 2026

Argentina's Patagonia is drawing attention as a possible site for large AI data centers. A cool climate, hydropower, wind energy, shale gas from the Vaca Muerta formation, and plenty of empty land are the draw, according to a Reuters report.

Power producer Pampa Energía plans a facility of up to 500 megawatts in Neuquén province and is looking for investors, with the first contracts expected by the end of 2026. Interested parties want to start with 20 to 40 megawatts. The grid infrastructure for the full build-out alone would cost around $900 million. Polish company Green Capital wants to start with 300 megawatts in Chubut, with 3,000 megawatts planned over the long term. OpenAI announced a project with Sur Energy back in October, but it hasn't been signed yet.

Unlike in the US, there's been no pushback against server farms in Argentina so far, because the country barely has any large facilities yet. That could change. Pampa wants to build next to its Loma de la Lata power plant, and Indigenous Mapuche families live there who have clashed with energy companies several times before. Grid connection, undersea cables, and how stable investment conditions will be after the 2027 presidential election also remain open questions.

Comment Source: Reuters

Maximilian Schreiner

Sep 8, 2026

Plus AI and society

ASML locks in TSMC, Samsung, and Intel while Huawei races to break its grip

Manuel Uth

Sep 8, 2026

Short News

Meta will no longer judge its engineers by how much they use AI tools. AI dashboards and token counters won't factor into performance reviews anymore, executives Maher Saba and Santosh Janardhan said in an internal memo seen by The Information. What counts now is the quality, speed, and complexity of the work.

Meta had previously made AI use a review criterion, which led to what people started calling "tokenmaxxing." Employees burned through AI tokens in bulk just to look good on internal leaderboards. The excessive use came with a price tag. Internal AI use alone is heading toward billions in costs in 2026, which is why Meta plans to roll out budgets and a central dashboard starting in 2027.

At the same time, Meta is testing its new AI agent tool Hatch, which is meant to handle computer tasks on its own, according to WIRED. But the tests are running into pushback inside the company. Some employees are reluctant to connect it to their personal accounts over privacy concerns.

Comment Source: The Information | WIRED

Jonathan Kemper

Sep 8, 2026

Short News

Three years after launch, Mistral AI has closed a 3 billion euro Series D round, pushing its valuation past 21 billion euros. According to Mistral, it's the largest equity funding round any European tech company has ever raised. Samsung Electronics is leading the round, with the Scaleup Europe Fund (managed by EQT) and PSG Equity joining as co-leads. New investors include Advent, funds managed by BlackRock, and the Grand Duchy of Luxembourg. Existing backers like a16z, Nvidia, ASML, and General Catalyst are also participating.

The company's valuation has roughly doubled. When ASML invested 1.3 billion euros in September 2025, Mistral was worth about 12 billion euros. In March, the company also took out an $830 million loan to fund its own data centers.

Mistral's models aren't top performers, though. Mistral Medium 3.5 trails Chinese competitors like Qwen and Kimi among open models and doesn't compete with closed US models at all. The company leans into enterprise instead. It operates in 20 countries and serves more than 125 companies, including Airbus, ASML, and HSBC. Since early 2026, Mistral has benefited heavily from European customers looking to cut their reliance on US providers.

Comment Source: Mistral AI

Matthias Bastian

Sep 7, 2026

Short News

Anthropic has signed compute contracts worth up to $517 billion in eleven months, according to The Information. Since October 2025, the company has locked in at least 14.8 gigawatts of computing power on top of the one to two gigawatts it already had and is planning its own data centers. Total planned capacity likely still falls short of OpenAI's 30-gigawatt target for 2030, though many of Anthropic's contracts extend well past that date, making a direct comparison tricky, The Information notes. Neither company can cover these commitments from revenue alone just yet. Anthropic's annualized revenue topped $65 billion, according to Bloomberg. OpenAI was above $40 billion as of July.

In early 2026, Anthropic CEO Dario Amodei warned against investing too fast, saying competitors "don't really understand the risks they're taking." Now Anthropic is the one racing to catch up. OpenAI CEO Sam Altman, meanwhile, is urging caution and warning about "unsustainable silliness" from neo-cloud providers. Technical progress, he argues, could turn today's expensive projects into bad bets.

Comment Source: The Information

Matthias Bastian

Sep 7, 2026

Short News

GPT-6 Astra has played through all of Portal on its own. Developer cozyblaze posted on X that the model navigated the full game and reached the credits without any human help after the initial goal was set. The run took about 23 hours and 43 minutes. Token usage adds up to at least $570 at Astra's list price, though cozyblaze used a $200 Codex subscription. The model controls Portal through MCP and a modified SourcePauseTool that pauses the game while it thinks. During each pause, the agent sees screenshots, player position, and camera angle, picks its inputs, then lets the game resume. The video below has those pauses cut out.

Code and docs are on GitHub. Cozyblaze noted that OpenAI set a goal in 2016 to solve many different games with a single agent. Problems remain, but watching an agent beat a full game on its own is a small taste of that original vision. GPT-6 Astra is "the worst model we'll ever get," cozyblaze added.

Comment Source: via X | Github

Matthias Bastian

Sep 7, 2026

Short News

Similarweb data shows ChatGPT still dominates AI chatbot website traffic and has recently regained share, climbing from 52.7 percent three months ago to 55.5 percent. Google Gemini, after a strong comeback, is slipping again from 27.8 to 25.6 percent. Year-over-year ChatGPT has lost significant ground (down from 73.3 percent), while Gemini doubled its share and Anthropic's Claude grew from 1.9 to 9.3 percent. DeepSeek (3.4 percent), Grok (2.4 percent), Copilot (1.6 percent), and Perplexity (0.9 percent) remain bit players.

ChatGPT has shed market share over the past twelve months but recently stabilized. Gemini and Claude are gaining ground, while the rest are stagnating. | Image: Similarweb

These figures cover website traffic only. Google likely drives many mobile Gemini interactions through its Android ecosystem that don't show up here, sending push notifications that open the Gemini app directly with a summary instead of a web link. OpenAI is also pushing its mobile ChatGPT app and the new desktop version ChatGPT Work, neither of which appear in the stats.

Comment Source: Similarweb via X

Matthias Bastian

Sep 7, 2026

AI research

AI-designed drug appears to turn back the body's biological clock in early trial

A study in Nature Biotechnology suggests that rentosertib, a drug designed with AI by Insilico Medicine, may reverse markers of biological aging. Six independent aging clocks predicted that treated patients were biologically up to six years younger than the placebo group. The trial covered just 42 patients and the drug hasn’t been tested in healthy people yet.

Read full article

Comment

Maximilian Schreiner

Sep 7, 2026

Short News

In Kenya, ChatGPT wiped out an entire business model: writing academic papers for foreign students. Thousands of Kenyans wrote essays and term papers for students at universities in the US and UK, in fields like medicine, computer science, and engineering, sometimes using their clients' university logins, the New York Times reports.

At its peak early this decade, at least 40,000 people worked in the industry in Nairobi alone, according to researchers. Teresios Bundi, 34, wrote more than 2,500 papers over twelve years and later charged $40 to $70 per text. After ChatGPT launched in 2022, prices and orders collapsed. Other work in the country also dried up, like transcription, data annotation, and content moderation for Meta.

Kenya's government had deliberately promoted online gig work starting in 2016, and companies like Samasource helped label AI data. About 80 percent of jobs in the country are informal. What's left, according to the NYT, are "humanizers" who rework AI text to slip past plagiarism checks. Oxford professor Mark Graham expects similar upheavals worldwide. Bundi warns, "A.I. is coming for bankers, for accountants, it's coming for engineers, and A.I. will come for architects. It's coming for everybody."

Comment Source: NYT

Maximilian Schreiner

Sep 7, 2026

Short News

Swiss banking giant UBS wants AI skills from its incoming junior bankers. Anyone applying to start in 2027 as a graduate or intern in Global Banking and Markets has to show how they can use AI to improve outcomes and efficiency, the Financial Times reports, citing people familiar with the matter. Interviews will include questions about how candidates use AI.

The requirement sits alongside classic criteria like a strong degree and will also apply to other newly posted roles. That makes UBS one of the first major banks with this demand. Banks are increasingly automating routine junior work like financial analysis, research, and client presentations. UBS said AI skills complement academic and social abilities but don't replace them.

Spain's Santander is also looking for advanced AI users for some trainee programs, according to its documents. Morgan Stanley analysts expect more than 200,000 banking jobs in Europe to disappear within five years. JPMorgan's Europe head Conor Hillery warned in December that juniors can't lose the fundamentals. UBS is also testing analyst avatars for client presentations.

Comment Source: FT