toponijo

6 ms·

1.
by toponijo 3mo ago
It can write correct code, but it's still really not good at that, at least without lots of prompting. In my experience, that prompting is usually just doing the design/planning work I'd do with or without AI.
2.
by toponijo 3mo ago
Given this, you can conclude that writers should be putting in at least at much effort as readers, whether or not they use an LLM. What really seems to be the problem is writers that don't at least check their own work, and pass that b
3.
by toponijo 3mo ago
I think it's just a tech thing! Same as when the internet really got going or what's happening with AI
4.
by toponijo 3mo ago
The point of going slowly is that we make sure something works, even if it has these bad side affects. Do we try experimental drugs with worse effects so that we can find effective ones faster? There are brave souls out there who will parti
5.
by toponijo 3mo ago
If you take these things together, they may be able to use the scanning feature to verify the existing stops. If it's non-existent no one could scan it, and they could remove those stops
6.
by toponijo 3mo ago
They did specifically say the safeguards are only more relaxed for those in their cyber program
7.
by toponijo 4mo ago
No sawdust means no tokens. Maybe there were already enough planks around. Maybe a task was completed without AI. What matters is what was done and not whether there's sawdust / AI was used.
8.
by toponijo 4mo ago
"How can I shove AI into something" is just seeing if a new technology can be used to solve a problem that's already solved. It's low effort high value compared to solving a new problem. Although I agree that people are
9.
by toponijo 4mo ago
They talk about the possibility of a spill going into the environment, but if they know it might spill, can't they make it spill and capture it?
10.
by toponijo 4mo ago
Aside from the accessibility benefit mentioned in the article, I would think that using semantic HTML like this would also make AI tools more effective at understanding code
11.
by toponijo 4mo ago
I actually think it makes sense to hone models for coding and agentic capabilities. Those models will be specialized for those tasks, and the results will be cheaper and better. We can still have a general model and specialized models
12.
by toponijo 4mo ago
It also loves to add edge case handling where it's not needed and in poorly chosen places