iamflimflam1

Wait until you, your immediate family, or close friends get “unlucky” and then report back.

Man whose job depends on you not running local models doesn’t want you running local models…

At one point I was very hopeful that the progression from Infrastructure as a service to Platform as a service was going to simplify things. Instead we seem to have ended up with mind boggling complex services.

Writing some code after months of agentic coding is so nice. “I remember this - it’s fun!”

I made a smash up of the arcade games SmashTV and Gauntlet.

https://smash-gauntlet.atomic14.com/

You need to switch into a product manager/QA/UI,UX mindset.

Is what is being built the right thing? Does it work? Is it usable? Has it been built in an appropriate way? (Note this is not a code thing, it’s an architectural question)

Traditionally developers have outsourced these things to other people.

There’s a further thing you should be doing - if the questions above didn’t pass - why not? What went wrong in your process?

Don’t beat yourself up about it.

Weirdly, I was talking to someone earlier about this who had been deposed.

They ask repeated questions repetitive questions hoping that you slip up/contradict yourself.

He said it was an interest experience trying to repeat exactly the same answers he’d already given.

The ported original is here: https://golden-axe.atomic14.com/

And I tried to make an RPG version: https://golden-axe-rpg.atomic14.com

Some to think about with both Claude and skills.

Can you invert things. What’s deterministic and what needs “thinking”?

Can you create a script that runs things chaining together the deterministic parts with calling out to Claude for glue that needs “thinking”.

That would mean some google people going to jail...

This used to be standard onboarding practice everywhere I’ve worked.

Point new starter at the readme and get them to fix any issues (hopefully very few!).

The danger is, that you end up passing the decision model something where the choice is trivial. There's so much in the harness that you don't need any intelligence.

Interesting idea - so instead of "which move do you want to make?" it's "which location do you want to go to?".

The phrase “cutting your nose off to spite your face” springs to mind.

Nothing yet. Unfortunately it sometimes feels like our industry has been overrun by grifters and chancers.

I’m sure this has been a gradual and long decline. Maybe it even started with the dot com boom and accelerated with crypto. With AI it seems to have got worse.

The number of - “I did/invented Jev last year”, or, “here’s a version of Jev I vibed up last night” is getting a bit ridiculous.

Especially ridiculous is how the hacker news crowd seems to be taking these at face value…

There was one the other day with a compelling demo. But when you looked closely at it, it was feeding in the options with the word “best” on the option to pick and a fine tuned model designed to recognise that word…

We’re at the point now where this has to be a problem for the future.

They really don't match the performance of Jev. You can probably fine tune them to work well for a particular use case (assuming you have sufficient data).

But Jev is works pretty much out of the box without any fine tuning.

It’s frustrating that we can’t see the “thinking” - it’s like we only have access to half the conversation.

Probably important to call out this part of the post:

Zero-shot vs. Fine-tuning: Out-of-the-box base models score ~0.35 on the typed-decisions benchmark (near random). The 0.766 score is achieved by fine-tuning on the benchmark's train split. Treat Laya as a fast foundation model to specialize, not as an omniscient zero-shot oracle.