Andrés Testi

No soy abogado, ni hincha de ningún equipo, ni padre de nadie.

[Andrés Testi @andrestesti]

@weekendflight @parenlamanok Hola Giralt, como andas?

[Andrés Testi @andrestesti]

@Mehackearon2023 @fernandezpablo También inventaron las olimpiadas, la FIFA, la pelota y el futbol.

[Andrés Testi @andrestesti]

@PolentelmeGoat @DevoxParodia Y ahora con la variante "Messi tiembla, que viene cuccurella". Es lo peor que nos podría pasar como especie.

[Andrés Testi @andrestesti]

@digaormf @KalshiFC NO

[Andrés Testi @andrestesti]

@BacOnfries77 @frabigol @QuotesforGoal Seguro?

[Andrés Testi @andrestesti]

@JoaqunPrezRios2 @Rhevolver Y?

[Andrés Testi @andrestesti]

@DevoxPosting Y NENE?

[Andrés Testi @andrestesti]

@posta_sports Y kid abajo?

[Andrés Testi @andrestesti]

@kiketinez @laStacyborracha Y?

[Andrés Testi @andrestesti]

@tobillosdepapel Estos festejan los goles de Harry Kane.

[Andrés Testi @andrestesti]

@DevoxPosting

[Andrés Testi @andrestesti]

@FPaulsocial @laderechadiario Si juega con Argentina se activa el agente P.

[Ben Cera @Bencera]

Uber had a $500M Anthropic bill in a single month.

Mine was over $1M.

So I rented GPUs and now use open-source models.

100x cheaper.

The future of AI is Open-source.

[Andrés Testi @andrestesti]

@worldwakeup44 @Vegettismo9 @blxyer10 @SamzaTalksFooty @WinnaFC How many world cups have you won?

[Mitchell Hashimoto @mitchellh]

The problem with the "if it works who cares what the code looks like" mindset for agentic work is that it assumes the agent has a perfect understanding of "works." Realistically, things are underspecified, agents make bad assumptions, etc.

To be fair, agents are pretty good at unit test coverage. They're pretty bad at designing human experiences (API, CLI flags, etc.), especially cohesive ones for future roadmap plans they may not have visibility into (unless your backlog is perfect and vision fully laid out, which I doubt). They're bad at knowing where performance matters and what type (CPU vs memory tradeoffs). They're bad at where compatibility matters and where it doesn't (and tend to err on the side of preserving it without further guidance). Etc.

Unless you have this ALL specified, you can't possibly claim "it works" without taking a look and thinking about it.

[Andrés Testi @andrestesti]

@EunucoBalbuceos Los minicomics de Alcalá preceden a Filmation, el He-man bárbaro sí existe y fue el primero.

[Andrés Testi @andrestesti]

@rauchg @cramforce Boilerplate has verfication and maintainance costs. It does not matter if you generated it with AI. compact, expressive and reusable APIs are more valuable than ever. Less is more.

[Andrés Testi @andrestesti]

@dosinaga2 Me encantan estas teorías que se arman ignoramdo lo más básico de la ley de oferta y demanda.

[Mitchell Hashimoto @mitchellh]

I've got an agent in a loop optimizing a renderer with the goal to minimize frame times (and tests to measure). It got times down from 88ms to 2ms and allocations down from ~150K to 500. Sounds good, right? Wrong. This is exactly why agent psychosis is a big fucking problem.

As an experiment, I rewrote the Ghostty core render state in Go, with access to identically laid out data structures as Ghostty and the exact same validation tests. I made a purposely naive renderer (simple, correct, but slow). 88ms per frame with 150,000 allocations (horrendous, lol)!

I then kickstarted a Ralph loop to bring the frame times down. I told it it can't modify input data structures or the public API or tests (they're correct), but it can do anything else it wants. It got to work.

It has worked for about 4 hours. I've spent around $350 on this experiment so far. The results?

88ms => 1.5ms
150K allocs => ~500 allocs

Incredible right? Nope.

My hand-written renderer I ported has frame times (same benchmark) of ~20us (0.020ms) and 0 allocations in the update path.

This is the problem with psychosis and lacking systems understanding. If you don't understand the system, you're going to accept that this is an incredible result. If you understand the system, you'll see better solutions immediately and can do roughly 75x better on throughput.

The people who blindly trust agent output are in the former camp. They're sheeple, overdrinking from a fountain of mediocrity.

Standard disclaimer: I use AI all the time. I like AI. The point I'm making is to not blindly accept results. Think. Analyze. Learn.

[Andrés Testi @andrestesti]

@rauchg It seems that AI finally would not kill SaaS by the end of the year.