Chance-Device

The average consumer was not the kind of connoisseur that valued quality photographs. The average consumer wanted something cheap and simple that allowed them to take photos and skip bringing them in to the shop to have them developed. The average consumer wanted to have their photos on their computer without scanning them, and didn’t want to pay a Polaroid premium for instant shots.

Over time digital cameras got better and better until the technology was subsumed into phones. Few people are unhappy about how it went.

This is how these things go. An inferior good is made, competes with the incumbent on cost and convenience, and is improved over time to become the dominant kind of good. This is the definition of disruption. The software industry, after decades of doing the disrupting, is now being disrupted.

The consumer will be happy about it.

If you want to see paper retractions, you ain’t seen nothing yet.

> things that I do and that my colleagues do, this kind of like curiosity-driven, you know, applied math, computational physics type research has always been justified by, I would argue, intentionally blurring the line between what I would call, you know, science as product versus science as process

> science as product is very kind of clear-cut. It’s, you know, things like, you know, cure cancer, solve nuclear fusion, generate, you know, clean energy.

> And then there’s science as process, which is kind of the curiosity-driven stuff about, you know, like, “I want to understand protein folding,” or, “I want to understand, you know, turbulence,” or, “I want to understand quantum gravity,” or something. And broadly speaking, we have tended to justify the latter by kind of laundering it through the former

And the examples he gives are actually the more defensible ones, he talks about a friend of his working on some abstract algebra under the false guise of cryptography research later.

And it’s not just him saying it, this is simply true. He should be lauded for admitting it publicly, this is the only way any progress is made. At least, it used to be. Now it’ll be AI instead.

So as measured by utility, I absolutely believe we don’t need humans doing science into the future. I’m sure people will continue doing it, but not for utility, for enjoyment - as a hobby. Probably we’ll all end up as dedicated hobbyists.

However, with AI it actually may become so cheap that the scattergun random approach becomes more viable rather than less. It’s when human time and resources are scarce that you need to optimise. The hobbyist approach may therefore ironically continue, but without the hobbyists.

Take a look at this interview from two days ago: https://m.youtube.com/watch?v=oQypVVv1u1o

The interviewee is worried about the future of math research. He is not strictly worried about being replaced, instead he is worried that he will no longer be able to launder math-as-a-hobby through math-as-something-useful as is the case today. He lays out very clearly that grant proposals claim to have useful outcomes while the proposers know those claims are nonsense.

Business as usual in math, and frankly in all the other sciences, is to do research that furthers the researchers careers or personal interests and pretend that it’s somehow useful. This would be absolutely fine if it were privately funded, but it’s not, this is public money.

In every other endeavour, lying in order to get money is considered fraud.

We have collectively wasted a huge amount of taxpayer money and human time, entire careers, on things not likely to ever matter to anyone.

I look forward to science becoming automated so that we can have real progress instead of the current broken system.

No, and you probably wouldn’t want it if you could. You don’t know what mutations are lurking in the genes that have been switched off or the consequences of turning them back on.

I’ve thought about this a bit, and what resetting the methylation map really requires, in addition to actually resetting that map, is resetting your genome to a known good state.

That’s pretty radical, but to selectively switch methylation on and off with precision you’re already talking about science beyond anything we have, featuring custom engineered constellations of proteins that unwrap your chromatin from around histones and then walk along it changing methylation state, then pack everything back.

And doing it for each cell type, without applying the wrong map anywhere.

And knowing what all the cell types and their correct methylation maps actually are.

And assuming that methylation isn’t used anywhere, for example in the brain, in a way we don’t expect.

And methylation isn’t the only kind of epigenetic modification.

So, if we’ve solved all of those problems, rebuilding your DNA to a clean, mutation free state at the same time should be a breeze.

We really do need AI for this if any of it is going to be remotely feasible.

Yes, and proteinopathies, and likely lots else besides. We probably don’t even understand the failure modes that aging protects against.

What we think of as aging then becomes primarily the epigenetic response to the problem of DNA damage, much like the symptoms of a virus are mostly the result of the immune response rather than the virus’s direct effects.

You have a very institutionalised idea of what publishing means.

For all the downsides of losing the business of government contractors, it really would hold Anthropic to their stated values: they’ll never develop AI controlled weapons on this blacklist.

> At a press conference in Sydney, Marles said the incident itself was “relatively minor” and that it appeared no personal health information had been accessed.

https://www.theguardian.com/technology/2026/sep/24/openai-ag...

> At a press conference in Sydney, Marles said the incident itself was “relatively minor” and that it appeared no personal health information had been accessed.