softwaredoug

You have to make a more rigorous claim for pessimism IMO. People usually hyperfocus on one bad area then get very down. And we all have a strong negativity bias. It’s not rational There are many posit…

While the US stays oil rich, we should expect the US to be a laggard in ending use of fossil fuels. China and EU are not oil rich, they’ll make a faster transition. Of course it’s idiotic to actively …

The only small catch is that NSF did not acknowledge it was illegally withholding funds Congress authorized for the array. And so who knows what might change in the future. Still Congress seemed willi…

The context behind this is head of OMB Russell Vought thinks impoundments - not paying out congressionally authorized funds - is constitutional. So this seems like a legitimate retreat on one of many…

The real issue here is impoundments. Russel Vought thinks the President has the power to not spend funds allocated by Congress. They literally are acting as if the presidents budget request for NASA w…

You’ll feel isolated from peers. You might not have an environment conducive to work. Poor work/life boundaries.

The promoters of AI themselves seem pretty convinced of the negative impact to society. It seems to even be part of the marketing. There’s very clearly a Substack of putting everyone out of work, redu…

AI isn’t actually a description of consumer value. It’s a tool to create that value Selling an “AI” product is like describing a C++ compiler as a feature to someone buying a video game

Well probably the whole idea from Meta was to invest these engineers in growth areas. And they haven’t succeeded in that.

Especially as inference gets cheaper, open models proliferate, and it all just becomes ubiquitous and commoditized. You can’t keep this genie in its bottle for long.

Nothing being talked about with Mythos wasn’t a known AI risk 12 months ago. Those rules could have been established to guide frontier labs. But yes crazy things happen. Maybe it won’t catch everythin…

It doesn’t really matter what party does it The ideal case is a statutory agency with regulatory authority that sets very clear standards for what model capabilities can and cannot release. Those are …

That’s a terrible way to create AI regulations If they actually cared about this issue we’d have predictable laws and regulatory bodies that let companies actually plan There’s a reason royal fiat doe…

The problem is it’s arbitrary executive action It’s not a question of regulate or not regulate. It’s whether it’s ok for the president to not like your company and screw with you. It’s fun to point ou…

No this is bad What you want is a neutral regulatory framework that markets can plan against. Not random executive action that can easily be abused.

The administration should not be able to arbitrarily punish companies they don’t like. Full stop. We need a neutral regulatory body for AI with objective, predictable standards. Not random bans based …

The US already has export controls for model weights. It appears this sets a precedent for even API usage being restricted. That’s seems like an attempt at a broad precedent setting power grab for the…

> We reviewed a demonstration of this specific technique being used to identify a small number of previously known, minor vulnerabilities. These vulnerabilities all appear relatively simple, and we…

It's not just about being annoyed - it's very practical. If the creator exerted human effort, they'll be able to maintain it. They can take responsibility for it. Even if the value of t…

The research shows sharpest declines at 13 years. We have a lot of infrastructure to support young readers. But not the the older readers. It’s not just screens. Teenagers are overscheduled compared …

The paradigm used to be create good enough abstractions you can express what you need in a few dozen lines or whatever. Those lines will be clearer and more precise than English for describing what it…

It’s not really about replacing software engineers. But about commodifying it. More software engineers (or roles responsible for code) that work for lower pay might be the trend. Or to maintain a high…

I'm not literally saying you were social engineered. I'm saying all the incentives are there for you to organize your content. Incentives to make things findable is more important to search …

In my research grep is fine if you don’t care about tokens and you have less than 100k files. The direct corpus interaction paper [1] shows a breakdown past this level. In my personal experience you g…

I’m not anti data center, but Colossus 1 is the posterchild for the anti data center crowd. The grievances with Colossus 1 might be a major cause why other data centers have such high opposition. Poll…

> compute from its Colossus 1 data center near Memphis, Tennessee, that xAI — now part of SpaceX — originally built for its own artificial intelligence efforts. Is this the same Memphis data center…