wild_egg

I don't disagree but

> Will we converge on a bug free state?

I have never witnessed this state in my two decades of software work. Why would we think AI will magically achieve this?

Aperture supports extensions? What sort are you working on?

> Depending on where you draw the line for millennials the school computers could have been running anything from Windows 95 to Windows XP.

Weird. I distinctly remember Commodore 64, DOS, and Windows 3.1 at school. Kind of hard to use Windows 95 before 1995, but I am pretty sure I had a Windows 98 SE machine at home when my school finally got Windows 95.

Why is BASIC a weird choice? It wasn't really a choice at all. All the computers in my school ran it until I was 8 or 9 and the school board could afford to upgrade to Windows machines.

Ask a color blind person to spell "red" and "green", then ask them how many red apples are on the table.

LLMs work on tokens. They literally cannot see individual letters. Yet, unlike humans, they can instantly create a tool to solve this task for them and never be stumped by it again.

I'm so tired of this brain-dead gotcha comment.

Last time I did anything with a BERT, you had to train or fine-tune. Is that not still true?

For me the cool bit is that it's all in-context learning or whatever so you can use it in any domain with zero setup.

Maybe bert and co. could do all the same things before, but the way in which you use them is quite different and that helps a lot.

What's the point of switching to switch? I've only ever used checkout and not sure what I am missing out on.

As an old person, what's the modern attitude?

It's a limit on input tokens. So that's 3 50k requests per minute. At Cerebras speeds, that's about 5 seconds of usage per minute.

I was very excited last year for their coding plan but seeing a burst of requests pulse and then sitting there watching the cooldown reset is really not a great time.

Even though each individual request was fast, the sessions were only maybe 10% faster on wall clock time since there was so much waiting time.

Don't all the major ones have free tiers now? Pretty sure you can do this direct from the Google results pages even

Moore's Law has nothing to do with processors getting faster. Dennard scaling stopped working but Moore just slowed somewhat, not stopped.

Wait, you actually found a viable counter to The Bitter Lesson? Please say more

Why are you talking about markdown? LLMs output bytes and llama.cpp has to process those bytes correctly. What if the LLM outputs a null byte and causes a buffer overflow somewhere and starts writing into executable memory?

Again, we are not talking about agents or your Python API script but instead talking about exploitable flaws within the inference engine itself. It wouldn't output `rm -rf`. It would output literal CPU instructions that llama.cpp would start executing directly. The payload would never get back to your Python script.

Article isn't about agents. It's about the inference engine itself being exploited by a malicious LLM output before it is ever sent to your machine or harness.