FuckButtons

When mythos was released - I guesstimated that we had 6 months until hackers had similar enough capabilities that we would start to see real problems in the wild. Unless there is some change in the trend line things are about to get really quite spectacular.

The rate at which software is produced has never been higher, nor has the rate at which it is thrown away for being broken and useless. It’s not at all obvious to me that if you control for both of those rate changes that the amount of useful software with a lifespan measured in years created per developer (ie, the software that actually has long term value) has actually changed all that much.

No, because these are not generated using any model available on their api.

Is this closed source?

Yeah I’m the exact opposite, thought is almost entirely auditory.

It doesn’t have a stable abi, how could it be easier? I would argue that the vast majority of what rust gives you comes at a direct cost of precisely the ergonomics of manual memory management. It’s a tradeoff that makes sense much of the time, but it’s still very obviously a tradeoff.

While this is true - and I see the same sentiment as i am about to make ad nauseam - it is kind of amazing how far the goal posts have moved in such a short space of time.

Not sure why ds cache costs are confusing? Their kv cache is absurdly compressed compared to almost everyone else.

if you think about that statement a bit more deeply, how else are you going to incentivize a society as a whole to take risks? If all the risk from your actions is ultimately your responsibility, no one would take risks, no one would explore for the sake of it.

That’s certainly the long term goal, but it seems foolish to abandon it without a solution to the problem.

>(I'd need a Lisp that runs in the browser, for starters!)

We would be in a very different world if Netscape’s exec team had let Brendan Eich just use scheme as he had planned to.

You could say the same about more or less any language feature - the thing is though that having a standard feature people can agree on is better than several people’s divergent opinions on which macros are sufficiently aesthetically pleasing.

You’re only considering the front half of the network though, classification is a helpful example for explaining the concept - but if it follows through to the output then you could quite easily imagine just shunting some activation from one branch of the representation tree to another to steer behavior. Which we already do with steering vectors - but if you could read off the actual topology you could do that with much higher fidelity.

So make one. Not everything is for you.

presumably if you are generating text for yourself to read prior to you writing something for a human.

>I see this quote being thrown around a lot as if it supports the article's point, but this quote is just some of the most famously smart, skilled, and effortful people ever admitting even they cannot condense their knowledge that much.

But that’s explicitly not what the quote says - it’s not that they can’t, it’s that it takes work.

The question for them is similarly existential - who is going to pay for research grants when the machines do all the work? Why pay for a PhD student to work on some research topic if an llm already proved the theorem?

>The central problem is that the majority of people go into math because they love discovering mathematics and want to participate in the process. That itself is a huge part of the reward. That creates the community. Simply have huge numbers of people outputting math as if it were an industrial factory kills the joy in the process and so in itself destroys the community.

Seems like math is split in much the same way software is.

short term profit over long term thinking - because no one in the system is incentivized towards long term thinking. The counter balance should in theory sit within the remit of politics, but even senators can be bought and few are far sighted enough to realize the consequences of incremental deindustrialization.

It got 8% of mine.

It’s really bizarre to me that anyone would think otherwise. Though - I was aware of the need to spend on sufficient local llm hardware to buy what I could afford for that purpose in late 2024 - so maybe I’m an outlier.

Ah, hello me.

It’s also very useful signal that the reasoning trace is leading to solving open problems - you can be certain that you’re not landing somewhere inside the training data.

Other than the by the lean certificate you mean.

One step at a time - how reliant do you think the ai labs are likely to be on their own tools right now, today, let alone 1-5 years down the road?

Yes, other people exist who are not like you.

They increased the projected liability of pre-existing financial deals that include share options.

I have lived outside the uk now for the majority of the years of I have been able to drive.

The winner was a foregone conclusion, transport signage everywhere else I’ve lived and driven extensively other than the uk is a patchwork mess that has clearly had exactly no thought put into it.

It seems to me that the only realistic option is to have more exams, closed book, and then use gen ai yourself to help reduce the workload of grading. Possibly with some kind of curve assigning more weight later in the semester.

If you assume zero opportunity costs, but that’s a terrible assumption.