Ridiculous, uninformed comment. There is no question that Effective Altruism has done tremendous good overall by saving many, many lives. I say this as someone who disagrees with important aspects, such as valuing distant lives equally to local ones. They have also been more correct than anyone else about AI, pandemics, etc.
I don't think it's a coincidence that he is interested in non-LLM solutions, since he mentioned last year on Twitter that he doesn't have an internal monologue (I hope this is not taken as disparaging of him in any way). His criticisms of LLMs never made sense, and the success of reasoning models has shown him to be definitely wrong.
I'd recommend looking for other sources of information if you're relying on someone who co-authored the paper that introduced the most misleading and uninformed term of the LLM era: "stochastic parrot".
Nobody said that Germany's economy is collapsing, but it's certainly underperforming. It had a recession in 2023 (-0.3% GDP growth), stagnated with 0.1% in 2024, and is expected to grow by just 1.0% in 2025.
https://economy-finance.ec.europa.eu/economic-surveillance-e...
> The trains, despite all the ineptness of Deutsche Bahn, are generally pretty good!.
> Promoting Nazism is banned. It’s not a slippery slope, and the country hasn’t devolved into authoritarian groupthink. You just can’t publicly support Nazism.
"No national figures exist on the total number of people charged with online speech-related crimes. But in a review of German state records, The New York Times found more than 8,500 cases."
"After Mr. Grote later made remarks admonishing others for hosting parties during the pandemic, a Twitter user wrote: “Du bist so 1 Pimmel” (“You are such a penis”). Three months later, six police officers raided the house of the man who had posted the insult, looking for his electronic devices."
https://www.nytimes.com/2022/09/23/technology/germany-intern...
If we can match our existing intelligence (but it’s a jagged border of capabilities), our progress in creating superintelligence won’t matter because we won’t be the ones making it.
Old roulette wheels had large flaws, even in the 1960s: https://thehustle.co/professor-who-beat-roulette. 30 years earlier they must have been worse. So there is a chance he noticed some anomaly that he tried to exploit.
The blog post clarified it's about maintainers of critical packages, not all contributors. This could be limited to packages with just one or two maintainers, especially newer ones. And they could remain somewhat anonymous, providing their information to trusted third parties only. If some maintainers don’t accept even this, their commits could be put into some special queue that requires additional people to sign off on them before they get accepted downstream. It's not a complete fix, but it should help.
None of this matters compared to the number of apps available. Given the high price of the Vision Pro and the resulting low sales, it would make little business sense for app developers to invest in creating apps for it instead of for the Quest 3.