Copenjin
Copenjin on Hacker News
I don't know what happened at some point, they started benchmaxxing and using the model for real programming just became impossible, probably 3.1 Pro is still "hitting a snag oh boy" in loop since the last time I tried it for a new project from scratch.
The most stupid thing I've read this month, found the same deep technology knowledge in other interviews. This guy is not the CEO, btw.
What is included in the 98%? IOS on Catalyst and core routers too? Isn't 24 months enough for anyone to forget this stupid idea and the boy genius (who cares if he went through YC23, worthless nowadays)?
I wonder what the average Cisco's employee will think reading this interview, probably those with enough seniority will think back at the Chambers years and shed a tear (incredible, I know).
I have it on both web and agy, but you know when you wait for something and in the end it turns out to be not that good...
What the AA charts say is irrelevant to me.
Since 3 Pro we already know that Google's model are very good in benchmarks, but when you start programming with them... not really good and each model has its quirks that make me continue the project with something else, once I fixed gemini's bugs.
I mean "a human" as human classifier of the quality of a text, most wall of texts if not all I see retweeted ad nauseam online could be summarized in a few sentences. That's a lot of human hours being thrown down the drain for very little useful content.
There is no respect for the time of the readers. Nowadays it's not just the occasional post from yegge or patio anymore. In general, the first few sentences of an article should present a summary and convince me to go further to read the details.
It didn't really need Docker to "take off", I'm not even sure if it actually ever did the take off...
It's a language with its pros and cons, I like the simplicity, it's a good choice for services that just expose an API/GRPC/etc... and the performances are not that bad.
This "but rust is faster" obsession often comes from people that don't even know what they are talking about or are even able to consider if thinking about performance make sense for a specific use cases. Sometimes, or at least it did before LLMs, having something simple to modify was more important that chasing supposedly higher performance because the rust people keep blabbering about it.
> would Rust have been the lucky candidate?
Probably not, considering that the author give me a "simple solutions are best" vibe.
Need a human's thumbs up before reading a wall of text from lesswrong, sorry.
This obsession with Rust is very hard to understand
Is it finally time to open my unused SBCs drawer and dump them all on ebay?
I wonder how many here or elsewhere care about Hinton (or others) predictions, not many I hope.
Copying whatever from the options book by Hull is something I don't know how many people would have done.... I mean...
Just a random thought since it's related to popular terms/slang, I'm noticing that accusing someone of not employing "critical thinking" is becoming very common on social networks among those that probably/clearly does not even know what it means, not talking about you obviously. Personally I'm forcing myself to phrase that in some other way lately, I wonder if I'm the only one that noticed this trend.
Why? Serious question, they completely lack the context essential to understand, the previous generation does not seem to have an improved understanding either, they are just users. Consumers of ads.
Doesn't tell us anything. This would be good if that meant they've developed some kind of "taste" for human output, I highly doubt it and for sure it will not reduce over-reliance on "AI" for school homework. I will not be surprised when this will be considered a "lost generation" from many points of view.
I don't understand the need of need "Satire, mostly." at the bottom, are opinion on things so dangerous?
Didn't they say that they can support bigger models now?
Does something like this exists for the keyboards too? (have one of those in its box due to bad software...)
Couldn't muster the energy to read more than a few sentences :)
Nowhere, tagged #satire, the podcast does not exist too, nice name for a podcast though.
Ok, it's satire, check the tags.
Heroic effort, sifting through that code I mean, but frankly I would have started a new one from scratch, the only thing of value is the name/popularity of the original project.
Using an LLM to judge drawing giving a rating? Not the best idea.
I'm also puzzled by the fact that I expected EVERYONE to point this out but yours is the only comment about it.
Even if using LLM is the only way to do things at scale it does not mean that it's always the right tool.
It's the first time I see something like this then, interesting.
There are a lot of site that are republishing your show HN but this one is the only one that actually extended it...
Yeah, skipped most of them since they where clearly AI generated. I wonder if the authors expect that people will actually read their slop in the replies.
> Thoughts[^0] from Theo
I will stop here, sorry but I think we have limited time to listen to opinions and nowadays since they are abundant on social media we should give preference to the substantiated ones.
Horrible board, even at the old price. The Pi stopped making sense after the original one, if you exclude the Zero/Pico.