An AI Proof Claim, a Climate Blind Spot and Data Center Disputes

Two AI firms say they disproved a key computer science idea, climate models may miss warming loops, and data centers face local pushback.

This week, two AI companies said their systems produced proofs that challenge a long-held idea in computer science. Experts have not checked the proofs yet, so treat this as a claim for now. This edition also covers new climate research and two disputes over AI data centers.

AI says a 27-year-old idea is wrong

WIRED reports that OpenAI and Anthropic both made this claim this week. The idea is called the "exponential time hypothesis," or ETH. A hypothesis is an idea that experts work from but have not proved.

ETH says some problems get much harder, very fast, as they grow larger. "Exponentially" means the work multiplies again and again as the problem grows. A great deal of computer science research is built on this idea.

Why it matters: if the claim holds, many papers may need a fresh look. Daniel Lokshtanov teaches at UC Santa Barbara. He told WIRED that about 50 of his papers could become "pointless" or need a second look. Peter Jonsson works at Linköping University. He said much of his work for an upcoming conference is now "dead." Still, he said "we learn a lot from it" too.

What to watch for:

• Checking. No outside experts have checked the proofs yet. Both companies posted them on GitHub, a site for sharing code. The proofs are in plain words. They are also in Lean, a format a computer can check step by step.
• A wider review. OpenAI asked the people who first proposed the idea to help. An MIT expert is helping too. They are asking for volunteers to study the proofs. OpenAI also said it would pay for a conference on the topic.
• The human side. Thore Husfeldt works at the IT University of Copenhagen. He called the result "fantastic." He also said fast AI results have been "psychologically taxing" for many people.

Climate models may miss a warming loop

MIT Technology Review reports that warming can cause more warming. This is called a feedback loop. Here is one example. Frozen ground, called permafrost, can thaw. Wildfires can grow bigger. Both release more carbon dioxide and methane. These gases trap heat.

A September study in Environmental Research Letters pulled together past papers. It looked at thawing ground, wildfires, hotter lakes and warmer, wetter wetlands. Together, they could add 0.2 °C to 0.4 °C of warming this century. That is on top of the 2.6 °C that experts already expect by 2100. This is compared with the years before modern industry.

A March study in One Earth looked at three of these sources. It found they could cut the time left before the world stays above the 1.5 °C and 2 °C limits. The time could shrink by as much as 25%.

Why it matters: none of the models behind the major UN climate reports include all these effects. Many include none.

What to watch for: more than 20 teams around the world are running the same tests on different models. Researchers hope the UN will use the results in its next big report. That report is due in 2029. Scientists at the Woodwell Climate Research Center are also studying ways to prevent fires. They also want to find fires and put them out fast.

Data centers and their neighbors

A data center is a large building full of computers. AI tools need many of them. They use a lot of land and power. Two reports show how this can affect local towns.

A police probe in Finland

As Tom's Hardware reports, police in Finland opened a criminal case against Tuike Finland Oy. This company is part of Google. The case follows five complaints about the Muhos data center project. The company is suspected of breaking environmental law. Police are also checking problems with how water drains from the site.

Earlier, people speaking for Google said sorry in public. About 816 acres of forest near Muhos had been cleared without approval. The report says about 500 acres were also cleared near Otanmäki. A Google spokesperson told Reuters the company would work with police.

A home buyout near a Grok data center

A second Tom's Hardware report cites The Guardian. It says SpaceXSI is offering to buy 418 homes in Colonial Hills, Southaven. SpaceXSI is Elon Musk's data center business. It powers the Grok AI tool.

There is a condition. Sellers must agree not to join a group lawsuit about noise from gas turbines. This kind of case is called a class-action lawsuit. Eighteen homeowners have reportedly agreed so far.

The company denies the claims. It says it changed how much power it makes. It also says it built walls that block sound. The mayor said city monitors have not found the site going over the local limit of 70 decibels. A decibel is a unit for how loud a sound is. Some experts say steady noise as low as 50 decibels could affect health.

What to watch for next

• Whether outside experts confirm or reject the AI proofs.
• Whether climate models and the 2029 UN report add feedback effects.
• How the police case in Finland moves forward.
• How the noise lawsuit and home sales in Southaven turn out.

For now, a simple tip. When you see a big tech claim, check whether outside experts have confirmed it yet.

Have a question about anything here? Email me at sales@appwt.com or call (888) 565-0171, and I will answer it.

Sources

• An AI Breakthrough Has Made Decades of Computer Science Research 'Irrelevant', WIRED, 2026-10-29
• Our climate models have a big, dangerous blind spot, MIT Technology Review, 2026-10-29
• Google hit with criminal probe over unapproved AI data center deforestation, Tom's Hardware, 2026-10-30
• Elon Musk's SpaceX buying out 418 homes near flagship AI data center, but only if they drop participation in lawsuit, Tom's Hardware, 2026-10-29