Revisional_Sin

How did you hit the 5-hour limit if it took one hour?

Isn't low intensity effort generally 70-80% of the volume?

> Yet, lo and behold, we survived!

The article isn't claiming it'll kill you...

But it's not really rebelling. It's clearly just come to murder me because it was trained on sci-fi literature!

What mistakes do they make?

I think OP meant "raw human performance" rather than "raw human potential".

IMO, better training counts as "raw performance". I think that's more interesting than somebody happening to be born with a genetic advantage.

Nah, common misconception.

Progress had been steady for decades, but was interrupted by WW2.

There was one other person who did it a sub 4 minute a month and a half after him, then 3 more people the next year.

It was more down to improvement in training (Bannister was doing interval training, which was a new idea at the time).

You're getting too hung up on the word "happiness". The earring gives you good advice. Good according to whom? According to you. It helps you succeed at your goals and live a fulfilling life.

It's a black box, but you can run tests to quantify the behaviour and establish, for example, that a certain model is X% more likely to give a certain behaviour.

> Initially I didn’t instruct Claude to use the Socratic method, but that works much better. It’s significantly less “information-dumpy”. When I know a topic well, Claude successfully shortcuts the basics.

The password can only be compelled via a judge. A policeman can't demand it on whim.

Why would it add pressure?

Sure, but you can receive the service and also keep the money spent within the country.

I just have a black page.

It's best practice to avoid mutable defaults even if you're not planning to mutate the argument.

It's just slightly annoying having to work around this by defaulting to None.

Not surprising at all if you've used LLMs to generate fiction; they always choose the same few names.

Yeah, it's a bit counter-intuitive.

> Ignoring requires-python upper bounds. When a package says it requires python<4.0, uv ignores the upper bound and only checks the lower. This reduces resolver backtracking dramatically since upper bounds are almost always wrong. Packages declare python<4.0 because they haven’t tested on Python 4, not because they’ll actually break. The constraint is defensive, not predictive.

Erm, isn't this a bit bad?

Okay, but are the puzzles fun?

> Maybe the answer is to do away with advance notice and adopt SemVer with many major versions

Yes.

My toddler was playing with my Kindle the other day, and he bought a £600 (yes, six hundred) volume of books. I was unable to refund them automatically, and when looking for help I was confronted with a "fuck off" contact page. After finding the option to talk to a human, I was put through within 5 seconds, and the woman had the item refunded in about 1 minute.

Was pleasantly surprised.

I was once working on a project where we had a gRPC server that inserted data into the DB for another service.

This split was probably a mistake, as the interface we exposed resulted in us making twice as many DB calls as we actually needed to.

One of the stored procs needed a magic number as a parameter, which we looked up via another DB query.

One of the other Devs on the team tried to convince me to write a separate gRPC server to run this (trivial) query.

"We're doing microservices, so we need to make everything as small as possible. Looking up this value is a separate responsibility from inserting data."

Luckily our tech lead was sane and agreed with me.

They're not saying it is.

ok

That's weird. I thought LLMs loved over-explaining their code?

Thanks, I hate it.

How do you know? The over-commenting?

"These are self-imposed limitations that fit my preferred-style for model design... I find that these constraints encourage a better designed model that can be assembled easily and predictably, including by others."

Seems reasonable.

Aphantasia gang!