psadri

Whoever created this - you are the crazy kind of person I want to meet in real life.

There is a strange dynamic between data center build out and inference prices.

If the data centers do get built and come on line, we will see an increase in compute supply and that should lower inference prices.

If they don't, for whatever reason including financial ones, then we will not have enough compute to meet demand and inference prices will rise. But this would mean the financial concerns were misplaced.

Where is the equilibrium point?

Hmm. I use pico. They could use an agent to modernize it!

I was dreading this moment. RIP.

I’ll be re-watching The Inner Light in his honor tonight.

I had exactly the same thought around Apple’s “report spam”. I have marked hundreds of text messages as spam. There is enough training data in there for even a dumb classifier to filter some of this stuff out. But next day, there are half a dozen more, nearly identical spam msgs sitting unread in the inbox.

I have also found that replying STOP = sign of life, your number is marked as “someone alive on the other end” and gets resold to 20 new lists.

Perhaps the real answer is people’s buying power. If people can’t withhold their work, they could withhold their spending, ie boycott misbehaving entities.

I think that’s a different argument. Apple absolutely “copies” hugely successful, durable apps. Just read about “getting sherlocked”. The rest of the apps are transient, niche and it’s easier to just harvest your cut of ads/payments.

Luna is incredible for the price but I find 6 to be worse than 5.6.

This is a touching sentiment. Thanks for sharing it.

The problem with CSS is the cascade and how it is not scoped. It's like global variables in code. And that can be traced back to it's roots 20+ years ago as a way to style "documents" vs applications that are naturally composed of self-contained UI components.

They are better at comparisons between two items vs scoring it in isolation. You can then turn a set of pairwise comparisons into numeric scores.

Even though I started using LLMs as soon as they became available, I was until recently very cautious around using them autonomously. I was of the opinion that I needed to provide heavy supervision around coding tasks. Granted, that was probably justified based on the model capabilities at the time.

However, lately, I have had several experiences that have completely changed my mind - the here is a high level goal (involving gathering production logs, setting up an eval, running tests, iterating, including diagnosis and coming up with new ideas, etc. and don’t bother me until it’s done. The experience has been incredible.

This has been possible because the agent has a way to iterate and hillclimb against a metric it can measure. Relatively easy in the world of software.

I’m now convinced that we will unlock the same gains once make other domains similarly “iterable”.

Eg Agent comes up with 100 new proteins, runs in lab autonomously, gathers results, iterates. One month (or whatever) later, you have your new protein designed.

Write tests first. Have agent iterate until they are satisfied.

The point is that it boils down to writing the tests correctly, regardless of who is implementing the actual code. Hand-written code without test coverage has the same problems as AI generated code.

Now that the models are smarter / know more than most of us, a new bottleneck is finding out what you don't know. If you don't know about logistic classifiers and how they could be applied to your problem, you will not ask for it.

These days, I tend to start my coding sessions by the high level problem I'm trying to solve vs the prescriptive, specific solution I may have in mind. It often surfaces ideas and approaches that I did not know about.

The point was that you can go further than "write me a contact that says x,y and z". I actually believe the process I went through with the AI was a more custom, tailored experience than any interaction I have had with actual lawyers, who'd usually apply a template. The AI asked me a bunch of questions, many of which were relevant, and led to follow up questions. I could ask my own questions, make corrections, etc. and all along, they made sense to me. This part of the experience was great.

There could still be more unknowns that I did not encounter and by definition don't know about.

And the final output could be wrong, wording, exact legal terminology etc.

Can we say working with an actual lawyer is guaranteed to be better, more correct? Lawyers are people and they also make mistakes, wake up on the wrong side of the bed, hate their job etc, just like the rest of us.