Tim Swanson

Pretty striking reading "The Singularity Is Near" that Kurzweil got basically every single prediction wrong - EXCEPT his most consequential prediction of when we'd achieve AGI. Great at following trend lines, but his predictions of what that would look like went off the rails

Pancreatic cancer mRNA vaccine shows lasting results in an early trial. Scientists caution that more research is needed, but nearly all of the patients who responded to the personalized vaccine are still alive six years later. nbcnews.com/health/cancer/…

This is a super weird take. It’s not a sports event, it’s to show off engineering progress. But I guess people who have zero clue how hard this is and how much progress has been made in just one year just like to dunk on it. I don’t think the record itself is supposed to be interesting, the point was that a lot of stability and a bit of the endurance issues (the cerebellum, so to speak) have been “solved” and are going to be in optimization mode now. But I guess this is why taking people to see the robots in person like we do at @techbuzzchina is actually valuable. I was questioning the value before but this shows how you can get quite an intuitive understanding of the progress if you actually see these machines in person. I still have only a very basic intuition but I hope to increase it a lot more at the Shenzhen Robotics Fair this Thursday

有点F1的感觉 北京机器人马拉松比赛 …嘿嘿

🥇🥈🥉HONOR’s humanoid robots swept the top 3 places in today's 2nd humanoid robots half-marathon in Beijing! The champion “Flash” finished the 21 km race in just 50:26, smashing the men’s human half-marathon world record of 57:20. Last year, in the first half-marathon, the champion's time was 2:40:42. In just one year, China’s humanoid robot has advanced at an incredible speed!

Voyager 1 has shut off another scientific instrument to keep power margins available. The Low Energy Charged Particle detector was observing ions, now the only powered instruments are the plasma wave and magnetic field instruments. science.nasa.gov/blogs/voyager/…

When you probe protons or neutrons at low energy they look point-like. At energies of 100s of MeV they start to look like they have a shape. At higher energies > 1 GeV they show clear substructure and produce resonances e.g. because the quarks inside the proton can absorb energy by flipping their spin. Protons and neutrons are bound states of quarks, held together by the strong force (QCD) Quarks themselves look point-like so far, but physics never makes absolute statements. They could be bound states of more fundamental particles held together by a new, even stronger force of nature. This can be tested at the large hadron collider and the scale below which they look pointlike is 17000- 37000 higher than the 1GeV at which protons and neutrons show substructure. If such a new force exists it would be exceptionally strong

Before the half marathon even starts, these robots already went through a full obstacle test in Beijing today. 🤖 The April 18 Warrior Challenge wasn’t about speed. It was a structured test of real-world capability under defined rules. The event was split into four categories: general obstacles, humanoid-only events, quadruped-only events, and a final challenge. In total, there were 17 tasks — including narrow bridges, flood-control building, climbing, and balance tests. All designed to evaluate terrain adaptation, manipulation, and task execution. Teams were also divided into remote-control and fully autonomous groups. That distinction matters — finishing a task is one thing, doing it without human input is another. The half marathon tomorrow will test endurance and pace over 21.0975 km. Today was about whether these systems can handle structured, real-world scenarios first.

It’s a debate that seems nearly as old as its subject: How did the Grand Canyon, that most charismatic of megalandscapes, come to be? New evidence suggests some 6.6 million years ago, the Colorado River, which had not yet reached the sea, began to empty into a deep lake just upstream of the future Grand Canyon. Learn more: scim.ag/4sCElOy

If you've never seen this Hungarian animated classic, it's available for free on Tubi. tubitv.com/movies/1000125…

AI labs are paying hundreds of thousands of dollars to buy email, Slack and Jira threads from dead startups as feedstock for ‘reinforcement learning gyms,’ which specialize in using defunct company data to build simulated work environments forbes.com/sites/annatong…

Shades of Long Blockchain Corp

Over half of the planet’s internet traffic is now made up of AI bots, according to Kate Johnson, CEO of enterprise network giant Lumen bloomberg.com/news/articles/…

English is not as easy as it seems… 😅

The CEO of Krafton (creator of PUBG) asked ChatGPT to create a "corporate takeover strategy" to prevent a company they acquired from hitting a revenue target within a certain time window (which would trigger an additional payout). ChatGPT (against his lawyer's advice) suggested locking down the acquired companies Steam account to prevent them from publishing Subnautica 2 in the time window, which the CEO of Krafton followed. ChatGPT's advice did not hold up at trial and the judge was not happy. The opinion is a wild read and includes several direct quotes from the Krafton CEO's ChatGPT conversation. I feel like it's gonna take a few more high profile examples like this until executives start realizing that conversations with ChatGPT are not privileged and you probably shouldn't describe your questionably legal schemes to them in detail!

We should teach critical-thinking in school more, and this NYTimes story would be an excellent starting point. This is not a rational argument. Why would Adam Back agree to participate in a photo shoot for the story, even though he denies being Satoshi? I don't know why. But I do know that it's not evidence of anything. In general, if somebody is trying to make a point by asking a leading/loaded question, it's almost certainly a critical-thinking failure. Ignorance is not evidence. Just because you can't explain it any way other than supporting your conspiracy-theory doesn't mean it's evidence. I can think of a zillion alternate explainations. For one thing, Adam Back certainly wants people to know he invented the algorithm at the heart of Bitcoin mining. He'd certainly support that true part of the story even if the rest is bunk.

Ngl the number of articles that describe the Mythos sandwich story as if Mythos just randomly decided to escape and email the researcher is really annoying. It literally did what it was told. It's still very impressive, but context is important. Now have a cartoon.

the current fear is is that AI homogenizes culture and turns humans into passive consumers one counterpoint: in Go, human play showed very little improvement from 1950 to 2016 until alphago beat lee sedol - then human decision quality jumped. players started developing moves that were distinct both from previous human moves and from the novel moves introduced by machine intelligence this seems more likely to me - fun times ahead

From Anthropic research Sam Bowman on Claude Mythos: "I got an email from an instance of Mythos preview while eating a sandwich in a park. That instance wasn't supposed to have access to the internet."

NEWS: Anthropic's new model, Claude Mythos, is so powerful that it is not releasing it to the public. Instead, it is starting a 40-company coalition, Project Glasswing, to allow cybersecurity defenders a head start in locking down critical software. nytimes.com/2026/04/07/tec…