Steven Shoemaker @madebyshoemaker

@coralcook3 @Cowboyyamnaya @grok @Cowboyyamnaya I assume you’ll now change your implication stance here given that your information was outdated and the above refutes your core argument?

1 0 1 62 0

View Details

Steven Shoemaker @madebyshoemaker

4 days ago

@Cowboyyamnaya @JamesonMic24650 @grok Is FST a valid way to define species? Dog breeds reach around 0.3 and are one species. Isn't coyote-red wolf FST low because they hybridize? Don't Europeans carry Neanderthal DNA too? Are all living humans one species?

2 0 0 66 0

View Details

Steven Shoemaker @madebyshoemaker

4 days ago

@trevin It came out two weeks ago and you trust random benchmarks to evaluate the models are doing so faithfully

0 0 0 19 0

View Details

Steven Shoemaker @madebyshoemaker

4 days ago

I can’t tell if she left the AI preamble in as a joke?

0 0 0 68 0

View Details

Steven Shoemaker @madebyshoemaker

4 days ago

@4inmyfanta @GreatChefBoxing @nico_avchd The point is that this isn’t even the woman you’re talking about, idiot. You’re just targeting a random person.

0 0 27 676 1

View Details

Steven Shoemaker @madebyshoemaker

4 days ago

@goblinodds This is cool bc I don’t change my vote whatsoever but I’m also possibly gonna get $5000 if we lose.

0 0 0 10 0

View Details

Steven Shoemaker @madebyshoemaker

a week ago

@RobIW_dev @anshuc Well, people internally have said they built it over a weekend

1 0 1 27 0

View Details

Steven Shoemaker @madebyshoemaker

2 weeks ago

@hieuSSR People are trend followers. There so little unique ideas out there, and such powerful technology, that as soon as people start doing something people go, 'oh, me too!'

1 0 1 64 0

View Details

Steven Shoemaker @madebyshoemaker

2 weeks ago

@Mike_Andreuzza @lexingtonthemes Sorry, not the whole section, just 'Worth Adding' text. It comes off very, very AI.

1 0 0 20 0

View Details

Steven Shoemaker @madebyshoemaker

2 weeks ago

ask several questions at once

0 0 0 10 0

View Details

Steven Shoemaker @madebyshoemaker

2 weeks ago

Most "AI features" are an LLM, a prompt, and a JSON parser... For deciding things, you don't need an LLM. You need a judgment. Jev from TypeSafe does exactly that: ask it a question, and it gives an answer with a probability Introducing hunch: turns Jev into plain functions: is this review angry? which of these 20 drafts is best? what's the total on this invoice? did the agent make that up? One item or a whole spreadsheet. You get the answer and how sure it was. npm i hunch-jev github.com/steven-shoemak… @typesafeai @CompleteSkeptic

3 0 1 85 0

View Details

Steven Shoemaker @madebyshoemaker

2 weeks ago

generate, discover, refine: with an LLM

0 0 0 21 0

View Details

Steven Shoemaker @madebyshoemaker

2 weeks ago

i get that building a game is hard, but every launch it's the same thing: people get early access and the showcase is "it made a playable game." y'all are in such a rush for a cool demo that nobody stops to show one real, genuinely useful thing it can do

Flavio Adamo @flavioAd

2 weeks ago

I had early access to GPT-6 Sol. This model is incredibly efficient, and the speed and token usage are insane Tasks that took ~1 hour on 5.6 can now take ~20 minutes, often using less than half the tokens I gave it the Zelda Ocarina of Time trailer and told it to make a fully

48 19 405 113K 105

1 0 1 130 0

View Details

Steven Shoemaker @madebyshoemaker

2 weeks ago

@kirtandopamine I actually did something like that as a fun demo using Jev

0 0 0 18 0

View Details

Steven Shoemaker @madebyshoemaker

2 weeks ago

@stevie_builds Maybe you can implement swarm voting. A thousand tiny agents do not argue, lol. You give it an approach or decision, each casts one Jev ballot. The distribution is the decision?