seunosewa

5 ms·

31.
by seunosewa 4mo ago
Then smaller the models are, the longer they have to reason when dealing with complex problems. The trade-off is real.
32.
by seunosewa 5mo ago
Unless you log its mistakes and how they were solved in decisions.log
33.
by seunosewa 5mo ago
AI has completely ruined animal short videos on Youtube. Videos of pets behaving like humans are everywhere. At first they warm your heart, then you realize that you've been tricked.
34.
by seunosewa 5mo ago
What do you mean? Those are valid identifiers but programmers aren't required to use them.
35.
by seunosewa 5mo ago
Less verbose languages also use fewer tokens, saving precious context.
36.
by seunosewa 6mo ago
They dropped the output cost, butthe input cost is relatively high. This is a recent trend. Seen with DeepSeek 4 Pro as well.
37.
by seunosewa 6mo ago
API costs are really high compared to subs.
38.
by seunosewa 7mo ago
Now try to use it to develop a simple app.
39.
by seunosewa 7mo ago
We could have different Python package bundles: Python base. Python webdev. Python desktop.
40.
by seunosewa 7mo ago
requests should be in the Python standard library. Hard choices need to be made.
41.
by seunosewa 7mo ago
Or just change the reasoning levels.
42.
by seunosewa 7mo ago
Create it!
43.
by seunosewa 7mo ago
Successful vibe coders read code.
44.
by seunosewa 7mo ago
I'd suggest Go ahead of Rust. It's more accessible to contributors.
45.
by seunosewa 7mo ago
Antigravity has an autocomplete model too. Based on Windsurf's, I guess.
46.
by seunosewa 7mo ago
After such hard-won wins, ask the AI to save what it learned during the session to a MD file.
47.
by seunosewa 7mo ago
Large Java apps like Android Studio are not good at managing 8gb of RAM. Emulators are terrible as well. They don't play well with the swap feature.
48.
by seunosewa 7mo ago
Unless what you want to do isn't well represented in the training set.
49.
by seunosewa 8mo ago
Perhaps they can just make it an option??
50.
by seunosewa 8mo ago
How did you verify it?
51.
by seunosewa 8mo ago
Once you've gone through that, you might want to ask it to codify what it learned from you so you don't have to repeat it next time.
52.
by seunosewa 9mo ago
They introduced the low limit warning for Opus on claude.ai
53.
by seunosewa 9mo ago
Then I pass the review back to Claude Opus to implement it.
54.
by seunosewa 9mo ago
Most llms.txt are very similar to the compressed docs.
55.
by seunosewa 9mo ago
What if they used the same compressed documentation in the skill? That would be just fine too.
56.
by seunosewa 9mo ago
Or just reducing the reasoning tokens.
57.
by seunosewa 9mo ago
The degradation may be more significant within the day than at the same time every day.
58.
by seunosewa 9mo ago
Chatterbox-turbo is really good too. Has a version that uses Apple's gpu.
59.
by seunosewa 9mo ago
There's no reason why Claude can't use git to manage the folders that it controls.
60.
by seunosewa 9mo ago
How do you enforce a rule if you make it easy to bypass it?

More ›