HackerLangs
TopNewTrendsCommentsPastAskShowJobs

cognitiveinline

75 karmajoined 18 giorni fa

Submissions

28% Faster: The Blink Prototype vs. Apple's iOS Browser Engine

open-web-advocacy.org
2 points·by cognitiveinline·18 giorni fa·1 comments

comments

cognitiveinline
·22 minuti fa·discuss
Then that's a tirade against capitalism - don't direct it at AI. If you don't like the situation, push for socialist policies.
cognitiveinline
·24 minuti fa·discuss
I do enjoy a good stream of consciousness rant, so thanks for that.

A simple rejoinder - The OP is one solution to many households where they can just be "the human tutor". If you can't see that, you simply don't know the average household.
cognitiveinline
·6 ore fa·discuss
Apparently one where taxing you higher for better teachers does not elicit a "sure, yes".
cognitiveinline
·6 ore fa·discuss
They are proposing a solution. You're armchair commenting, at best.
cognitiveinline
·6 ore fa·discuss
Any model, including inexpensive ones. It's not a hard problem.
cognitiveinline
·11 ore fa·discuss
Manual knowledge of git to get out of tricky situations will be as passe as using log book for multiplication. It's just not required anymore.
cognitiveinline
·11 ore fa·discuss
> Give children books, creative toys, teach them read, enjoy music, do science or any activity always with a human companion

Privileged much?
cognitiveinline
·12 ore fa·discuss
ok. You've rephrased something. But do you have a specific solution? Because the OP seem to do.
cognitiveinline
·12 ore fa·discuss
You think aristrocrats are using GPT and succeeding?
cognitiveinline
·12 ore fa·discuss
Fine. And do you have a tractable solution? The Ello folks think they do, and are giving it a shot.
cognitiveinline
·12 ore fa·discuss
How? Taxing you higher?
cognitiveinline
·13 ore fa·discuss
Pray tell.
cognitiveinline
·21 ore fa·discuss
These agents follow instructions to never make a specific mistake, if you simply record it on context (like AGENTs.md).

You'll frustrate yourself by not using this tool without above. You'll definintely frustrate yourself expecting the tool to be "genuinely sorry"!
cognitiveinline
·l’altro ieri·discuss
Lol. Imagine you were called a stinky employee, and you thinking that's not a personal attack. Would be very noble of you.
cognitiveinline
·l’altro ieri·discuss
People are doing the simpler thing: moving ecosystems more suppotive of their goal.
cognitiveinline
·l’altro ieri·discuss
That's not the direction of LLM progress these past 2 years. Cost has always decreased per unit intelligence. Open weight models with lower parameters beat gpt4 that we were so impressed by.
cognitiveinline
·l’altro ieri·discuss
Humans can't become 100x efficient cost wise. That's how cheap LLMs are.
cognitiveinline
·l’altro ieri·discuss
Got a reference to something specific? Per the blog post the overall quality, speed and size all improved. And multiple users have corroborated.

Like have you run into a specific bug or seen a regression - that's the cause of your reaction?
cognitiveinline
·l’altro ieri·discuss
Everyone is calm. Many are just tired of you not taking a hint.
cognitiveinline
·l’altro ieri·discuss
Fable is kind of fantastic on the difficult tasks. if it's something eithe rmodel can do then you can't see the difference. Fable also makes much less mistakes. It's a more relentless, proactive problem solver.