strangescript

7 ms·

1.
by strangescript 11d ago
This is a step beyond baseless predictions. Tao also had a "weird" "hypothetical" comment about LLMs solving complex proofs with impossible to human verify Lean.
2.
by strangescript 3mo ago
The study is "new", but the data they are using is old. It has references GPT-4 for god sake.
3.
by strangescript 3mo ago
I understand the struggle, but I feel like so many of these projects are going to ban AI contributions right before it starts getting undeniably better than humans.
4.
by strangescript 3mo ago
The fundamental flaw in all these style arguments against core employment is people's personal desire to have a straight forward path for survival that doesn't involve a lot of strategic thought. Its hard for highly intelligent pe
5.
by strangescript 3mo ago
Agents are getting good but professing they are surpassing you in domain and architectural knowledge with no special prompting is basically self reporting at this point. That could be your job wasn't that complicated or your personal k
6.
by strangescript 4mo ago
Hopefully we never do something silly like making a lead pushing machine that operates at high velocity, then mass produce it, what a terrible precedence that would set.
7.
by strangescript 5mo ago
this is apples to oranges, the flash version version a full version this thing is like 5x better than flash at fine grain detail
8.
by strangescript 5mo ago
Anecdotal, of course, but the biggest change I ever made in my life was right before bed: take a screaming hot shower with dim lighting. I'd say 95% of the time, I get in bed and just pass out and have no real memory of time passing be
9.
by strangescript 5mo ago
Same reason why two companies have the same idea, one goes viral and one doesn't. Public opinion matters even if its illogical at times.
10.
by strangescript 5mo ago
Their concerns weren't completely off base, I think they just over estimated how much it would really matter in the grand scheme.
11.
by strangescript 6mo ago
I will "suffer" through .004 of electricity if I can run it on my own computer
12.
by strangescript 7mo ago
When you cannot win, you regulate the scoreboard. This is why Europe keeps getting lapped. They mistake control for competence. They cannot build platforms at scale, so they try to govern everyone else’s.
13.
by strangescript 7mo ago
If your co-workers are getting promotions and raises and you are not, its a you problem. If someone else is getting credit for your work, its a you problem. Given your claims of impeccable work, we are only left to assume its a personality
14.
by strangescript 8mo ago
If you read the PR, the bad issues are in a few extensions, not the bot itself. The unencrypted oAuth token isn't really a big deal. It should be fixed but its a "if this box is compromised" type thing. Given the nature of cl
15.
by strangescript 8mo ago
Same with food. Plenty of food, just not where its needed.
16.
by strangescript 8mo ago
I find gpt-oss 20b very benchmaxxed and as soon as a solution isn't clear it will hallucinate.
17.
by strangescript 8mo ago
Not really an apples to apples comparison. You are comparing it to core technologies that millions of things sit on. There will always be money for that.
18.
by strangescript 8mo ago
Its almost never correct to rip on a project from a distance. Only two things can happen, one, you are wrong, and the project succeeds. This is a personal catastrophe for your career at that company. Two, you are correct and the project fai
19.
by strangescript 8mo ago
This is cool, but once a week seems a little slow
20.
by strangescript 8mo ago
yes, AI isn't penetrating those fields with high job losses at all
21.
by strangescript 9mo ago
Cool project, but I never found the cloudflare DX desirable compared to self hosted alternatives. A plain old node server in a docker container was much easier to manage, use and is scalable. Cloudflare's system was just a hoop that yo
22.
by strangescript 9mo ago
this, provided you don't mind hopping around a lot, 5 20 dollar a month accounts will get you way more tokens typically, also good free models will show up from time to time on openrouter
23.
by strangescript 9mo ago
If your outlook is 10 years then for sure, its valid. I am not sure how you come to that conclusion logically though. At the beginning of the year we had 0 code agents. Now we have dozens, some are basically free, (of various degrees of qua
24.
by strangescript 9mo ago
Do they consider code readability, formatting and variable naming as "errors" for the overall count. That seems dubious given where we are headed. No one cares what a compiler or js minifier names its variables in its output. Yes,
25.
by strangescript 9mo ago
AI can make you a basic signal for whatever group you want with zero oversight now anyway. The days of trying to proxy anti-encryption laws so you can spy on your people are numbered.
26.
by strangescript 9mo ago
"This post is to underline that Nothing is inevitable." Inevitable and being a holdout are conceptually different and you can't expect society as a whole to care or respect your personal space with regards to it. They listed
27.
by strangescript 9mo ago
I am saying it doesn't matter. Driving in motor vehicles is dangerous, but we still do it. It would be safer to never drive anywhere. We have decided its an acceptable risk. Someone doesn't post one car accident on hacker news whe
28.
by strangescript 9mo ago
Your assumption is you are "in control", as does everyone right before they have an accident.
29.
by strangescript 9mo ago
you should probably avoid driving or riding in motor vehicles
30.
by strangescript 9mo ago
I work 60+ hours a week with Claude Code CLI, always run dangerously skip, coding on multiple repos, on a mac. This has never happened. Nothing remotely close has ever happened. I have been using CC since research preview. I would love to k

More ›