kolinko

it may depend on where you live and who you hang around with

Your fundamental math is wrong

Same here

Why would it be random? Perhaps we’re talking about two different things?

I’m talking about asking an agent at the end of the session how well the session went - what tools it found useful, what it found confusing and whatnot. Agents do gladly complain about doing pointless work or fighting with the harness :)

I have a similar system and it does wonders. One of the questions is also to ask for just freestyle opinion of the task and the replies are hilarious at times.

“I existed for one word. The entire point of my life was that my process boot side-effected a module reload. 'ok'."

“ "The overhead-to-work ratio was astronomical."

“the koleo purchase turned into a boss fight with an undismissable Vue modal and a WKWebView that kept eating its own cookies."

Are you talking from experience or from theory?

I have a reflections mechanism, which launches right after the session ends (when it’s still in cache, so it’s almost free) and it catches a ton of problems with tool/skill/system design, all the time. Low level stuff can be fixed almost automatically (e.g. agent funbled 10 turns bc tool’s docs were wrong), high level too but human gating is useful.

If you try doing it any other way you have to pay for input tokens, not cache read, so it’s 10x more expensive.

Good one! Also, a simple thing - when your service gets miscalled (wrong params either via cli or http) instantly show help/explaination of the proper param uses. Can save a few turns of agents trying to figure out what’s wrong. Helps humans as well :)

It’s also ironic that the slop article complaining about the lack of originality has very little intellectual originality itself.

What you’re saying here makes no sense.

Manifold vs space - yes, I meant manifold, english is not my first language and I missed this word.

The argument stands.

That is not true - a model trained in the internet can both build verifiers to remove false/poor quality data from the next training, and build synthetic datasets that will supplement its training.

Similar to a human that wants to learn something and invents exercises to practice.

1-2 years ago it was a theory, but new models are trained, successfully, on synthetic datasets.

There isn’t a big difference between interpolation and extrapolation when the space has an immense amount of dimensions, and when you are free to modify the space at will.

Why? Just get AI to grafe them and give you the outliers to judge

This is a recurring sentiment but flawed, I think.

First of all, neural nets do nit return averages per se. They construct space between the points and extrapolate outside of the points. So even if a point was not in their training data, they will be ok, in many situations, to acknowledge it.

Or in other words - LLMs don’t average. They construct world models. A novel thing that fits their world model will be accepted no prob. A thing that doesn’t may still be accepted but with challenges.

The same is true though for humans, including scientists. There is a saying that science moves one grave at a time - because often prev gen of scientists needs to die off for a new idea to take root.

Or in yet other words - even if llms produced averages, an average of a discontinuous set can lie outside of that set. And the set of all human ideas is very much discontinuous.

1M context - at least with most of the projects I ever worked with, 1M, or even 100k would be enough to explain in broad strokes the class/project/deployment structure, and a window of 200-500k to explain the specific issue at hand.

I can show them a finger.

And seriously - what about people without hands? What about scammers pretending to be Google gaining access to my camera? What about blind people? What about people using the site in places where camera use is not allowed?

From my benchmarks, sadly, it doesn't seem to be the case much. Surprisingly. I found Sonnet comparable in speed to Opus (sic), but perhaps I was testing it wrong?

The whole third part of the posted document is about relative urls.

It’s not that difficult, it’s just a system prompt and a set of basic file edit/bash/etc tools.

Me, personally, I didn’t build it from scratch but I ported original CC from published sources into Python and extended it to match my own requirements.

It is most probably metabolism/genetics dependent. For me, I didn't have much withdrawals (aside from a first week), but even a month later I was insanely sluggish and it was difficult for me to function properly.

The point of modern mefivine is that we have drugs that have benefits that outweigh risks and downsides.

And downsides of penicilin? There are some but they are massively offset by not having alt consequences

What reports?

“be more disciplined” - did you read any studies about the sources of obesity? Or do you struggle with weight since childhood yourself?

This assumes underlying behaviours are core reason. Research (tein, adoption) suggests otherwise

Just because it wasn’t hard for you doesn’t mean it isn’t hard for the others. Grom what you said it seems you just had some bad habits you had to fix and that’s it.

Not the size of function but conplexity.

It may depend on your specific workload. E.g. for regular webdev work Opus is more than adequate, for heavy duty data analysis, for experimental stuff and for complex systems it was night and day.

I had only a few places where I did spot a difference but that difference was significant and I can imagine where people would be amazed.

Ed Zitron is amazing at cherry picking data to fit his thesis.

What are you talking about?

Prices of lowest tiers of models have fallen how much - 10-100x over the last two years.

And actually, the model quality you needed to pay for in the past, you can just run on device now essentially for free.