karmasimida

6 ms·

61.
by karmasimida 7mo ago
Quite on the contrary for my experience. xhigh is the only model + thinking level that can reliably locate the bug
62.
by karmasimida 7mo ago
I have burnt billion of tokens in gpt 5.4 and I didn’t know what you are talking about
63.
by karmasimida 7mo ago
Just use GPT5.4, avoid the drama and it is a better model anyway
64.
by karmasimida 7mo ago
Those are bigger models. The serving isn’t going to be cheaper. Why expect cheaper then? The performance is also better
65.
by karmasimida 7mo ago
HN is in denial, which is understandable AI is already better at understanding code than 99.99% of human, the more I use it the more I believe this is true. It can draw connections between dots far quicker and accurate than a human could ev
66.
by karmasimida 7mo ago
This is true. When I am using codex, compaction isn’t something I fear, it feels like you save your gaming progress and move on. For Claude Code compaction feels disastrous, also much longer
67.
by karmasimida 7mo ago
Well I really don't like my handwriting, would rather avoid it
68.
by karmasimida 7mo ago
I mean you don’t need your first job go to top of the top companies. Your first job is to get you into the industry then you can flourish. How many juniors OpenAI GDM are going to hire in a year, probably double digits at max, the chances a
69.
by karmasimida 7mo ago
This is definitely the Claude killer OpenAI is cooking. And so far it has succeeded
70.
by karmasimida 7mo ago
You should respect the government’s choice. It is elected after all
71.
by karmasimida 7mo ago
As programmers become intelligently irrelevant in the whole picture, you would see more posts like this
72.
by karmasimida 7mo ago
It is a bigger model, confirmed
73.
by karmasimida 7mo ago
And he is back to Pete hegeseth now? Lollll
74.
by karmasimida 8mo ago
Raw scale of parameters is POWER, you can't get performance out of a small model from a much larger one.
75.
by karmasimida 8mo ago
He is a horrible public presenter. He presents himself as someone who is nervous, validation seeking, yet it is stupid of you to not trust him. He lacks confidence yet feels incredibly arrogant. He would succeed in academia as the principal
76.
by karmasimida 8mo ago
Sam is a player and honestly the more interesting one in the whole thing. Mad respect to Sam, now I believe OpenAI have better chance to win in the race
77.
by karmasimida 8mo ago
Dario is being ruled out due to ideological standing Makes perfect sense
78.
by karmasimida 8mo ago
It gets so much money, compute and US user data. It won’t be allowed to operate as is as a foreign entity Best scenario it will get TikTok-ed, otherwise it will become the real national security risk Had the exit happen, well, as US has a m
79.
by karmasimida 8mo ago
They won't, not Google at least There is a whole situation with dealing with military contract during Trump 1, and it didn't go well for Google, I doubt Sundar will go the same route once again.
80.
by karmasimida 8mo ago
Safely in what way? If you ask them to stop, the easy argument is Chinese won’t stop, so they won’t stop. Essentially they will not stop at all, because even they know no one can stop the competition from happening. So they ask more control
81.
by karmasimida 8mo ago
Precisely Anthropic never explains they are fear-mongering for the incoming mass scale job loss while being the one who is at the full front rushing to realize it. So make no mistake: it is absolutely a zero sum game between you and Anthrop
82.
by karmasimida 8mo ago
Label them as supply chain risk and move on. Enough of this drama already
83.
by karmasimida 8mo ago
Same, I also have both subscriptions (100 Max and 200 Pro), and I am considering canceling MAX plan but would give it another month on watch. The doomer sentiment is quite baffling to me, what trouble is OpenAI in? Definitely not after GPT
84.
by karmasimida 8mo ago
It would be an unpopular opinion but I am here to say it: OpenAI has the best model, that is how they are going to compete. Their chatbot business could be in trouble, but Gemini needs a LOT of work to make it better to use too. Coding wise
85.
by karmasimida 8mo ago
> Anthropic Claude has the best integrations with coding Disagree, Codex is neck and neck with A/ on coding front
86.
by karmasimida 8mo ago
Unless they stop selling APIs to the public this can’t be stopped. Mind you that nuclear weapons are able to be regulated not because the tech itself is secret, it is because the refining is nation state effort, that is impossible to go unn
87.
by karmasimida 8mo ago
True, but anything remotely useful is 300B and above
88.
by karmasimida 8mo ago
Does local AI have a future? The models are getting ridiculously big and any storage hardware is hoarded by few companies for next 2 years and nvidia has stopped making consumer GPU for this year. It seems to me there is no chance local ML
89.
by karmasimida 8mo ago
Accomplish the task I give to it without fighting me with it. I think this is classic precision/recall issue: the model needs to stay on task, but also infer what user might want but not explicitly stated. Gemini seems particularly bad
90.
by karmasimida 8mo ago
Gemini just doesn’t do even mildly well in agentic stuff and I don’t know why. OpenAI has mostly caught up with Claude in agentic stuff, but Google needs to be there and be there quickly

More ›