latentsea

7 ms·

121.
by latentsea 4mo ago
Exactly. You get it.
122.
by latentsea 4mo ago
>It's the reason the costs of things like education and healthcare are downright extortionate, the reason you're paying back your college well into your fifties, the reason you don't call an ambulance for someone in the US
123.
by latentsea 4mo ago
This is part of the training data now. She can hear you, you know...
124.
by latentsea 4mo ago
Cos I'm too busy doing it at work.
125.
by latentsea 4mo ago
> The first one-shot app was created with ChatGPT in June 2023 - 3 years ago. In my experience, the current result of one-shotting apps is just as bad today as it was back then. Hard disagree. Take a good SaaS starter template and do a b
126.
by latentsea 4mo ago
> I know of a few places that have gotten such large gains from LLMs that they know have their engineers working on creating homegrown ports of popular services (Docker etc.). Sounds like a good way to eventually erase those gains.
127.
by latentsea 4mo ago
Our whole team has transitioned to agentic engineering around the start of this year. We've been heavily investing in harness engineering to ensure final outputs are production quality and not slop. These days it very much feels like t
128.
by latentsea 4mo ago
LLMs learnt to be confidently wrong from us.
129.
by latentsea 4mo ago
They said that they hadn't "suffered" from hallucinations in months due to effort they put into harness engineering. It's not quite the same thing as saying hallucinations are never happening for them. I find hallucinati
130.
by latentsea 5mo ago
Send them to a country where college is much cheaper and then also be able to give them like $400k ~ $450k
131.
by latentsea 5mo ago
They kinda do though, that's sort of how agents work. At least that's how it's always felt to me.
132.
by latentsea 5mo ago
> That sort of thing seems likely to matter until machines can replace us completely. This seems like the crux of the issue. Like people are banking on that day coming even if they don't know exactly when.
133.
by latentsea 5mo ago
Yeah CPAP is doing the heavy lifting most likely.
134.
by latentsea 5mo ago
Humans hallucinate because they take shrooms or have schizophrenia.
135.
by latentsea 5mo ago
Wait until you try d4vid, it's killer.
136.
by latentsea 5mo ago
I'll bet you "neon" is somewhere in those lyrics.
137.
by latentsea 5mo ago
YouTube is excellent. Aside from my main account I have one that I mainly just use to listen to music, and I just surf the algorithm listening to whatever is in the recommend list, which is usually a handful of songs I've got on heavy
138.
by latentsea 5mo ago
> There is personal expression in the lyrics. Sometimes there is. Sometimes it's just made up shit with no real backing in the same way a lot of stories told in comedy routines are. Not all of it is genuine expression. I love music
139.
by latentsea 5mo ago
Well, auto-compaction is a thing in Claude Code now. Plus we have /goal command and some automated review stuff, so you can kinda just get it to loop until the automated reviews are satisfied and CI is passing. Does most of the heavy l
140.
by latentsea 5mo ago
https://openai.com/index/harness-engineering/ https://www.anthropic.com/engineering/harness-design-long-ru... https://www.anthropic.com/engineering/effective-harnesses-fo
141.
by latentsea 5mo ago
Well he doesn't write CRUD apps, which plenty of us do, and with a decent harness, agents can do decent enough work on them.
142.
by latentsea 5mo ago
This shouldn't actually change virtually anything. We had this happen recently, and were able to rollback within minutes. Devs hand-coding stuff breaks things too. If you already have good observability, fast rollback processes, and fe
143.
by latentsea 5mo ago
> Yes, I don't have anything important to say other than I 100% agree with this comment. AI in its current state is akin to Stack Overflow and Google on steroids, but from my experience, it doesn't do well building out full-sca
144.
by latentsea 5mo ago
Waiting for them to come down any day now. Been waiting since 2017.
145.
by latentsea 5mo ago
I was impressed that I was able to take the same basic idea and apply it to anything that a Claude could construct a metric for. It's nice being able to just run /autoresearch and speed up your test suites, and shave time off your
146.
by latentsea 5mo ago
Well that's not my problem, that's yours.
147.
by latentsea 5mo ago
Not even remotely the same situation.
148.
by latentsea 5mo ago
Tell me one for that specifically, or gtfo.
149.
by latentsea 5mo ago
imagine the horror when after practicing the skill diligently for years this cognitive load never reduced beyond 2x ~ 3x that of your native language.
150.
by latentsea 5mo ago
Crime has always payed if you do it right... news at 11?

More ›