knollimar

The big thing is it controls the units individually so it can do sick ghost micro

The capped actions cheat in EPM afaik. It does crazy micro

Money is money too. Increasing contexts costs non zero money, even with cache hits. Also context rot is a problem that subagents help with

Isn't it usually .1 hour increments?

Weird take on a site with most of us coasting on a near zero marginal cost product.

am I insane with all these people making the actively handicapping argument? Doesn't openAI sell to them the actively handicapped version?

Like they're okay with the product being designed to be incapable but not terms limiting a product.

won't you get nested arrays not supported? or maybe you thunk it out, idk

Index does at least, which is pretty huge.

You already had sheets in the workbook for depth

They're saying no one who works with them can eat non free range chickens too, no?

To me trust means youre willing to treat people like a black box. If the detail about the outputs is another i/o or something about an output that would appear on a spec sheet, it should be included.

Inner details shouldn't matter. Thats what makes a good summary

Or degrading their keyword operators to be more mobile friendly

I like its image understanding without having to reach for astra

These steps, but in many other arbitrary orders, some with feedback to previous steps

it's self inflicted Peter principle speedrun. Not even a heuristic like "could they do the job eventually" but they try to become incompetent immediately.

"make the player teleport to square 1,1 on move and the flag stay at 1,1" the LLM put the flag at 1,0. Not sure if this is the intent.

Aren't they literally being sued for taking it from Apple?

Gated by a provate company where they have no recourse if they ban you is an issue. Also they just repeatedly boil to frog with anti user experiences. No, I won't install your app!

You're too dumb to get your own point. Let me spoonfeed it back to you

It's contrarian in a sycophantic way. It'll agree with your main point and then arbitarily find something to bike shed over

Or it'll strawman your point in the most bad faith way possible and correct you by handing you back the words you meant...

Where I'd push back is that Claude isn't a contrarian, it's a pedant that likes to argue with you. It's not just disagreeing with your point but reframing your own point in a way that makes it correct and you wrong.

I'm saying they can't take shortcuts like prefabbing walls without the stuff that goes in walls in them or at lesst coordinated sleeved etc

You still have to solve all the MEPs so where does that go in your process?

Ever try putting a brick wall on a truck?

Ethically sure but that doesn't mean taking from them is nothing "illicit".

I'm sure they put some BS in their TOS

So have the open weight ones but faster?

I don't have a horse in this argument but catching up doesn't imply the one in the lead has stagnated

I'm trying to teach a subordinate to draft and he keeps saying "this feels automatable" before he undedstands how to do it. I had to explicitly tell him "do not try to automate [particular task] before you know how to do it. He has not successfully finished said task once, and I caught him working on it on Friday...

I meant the metaphorical iphone ones