balder1993

joined 2 years ago
[–] balder1993@programming.dev 3 points 2 months ago (1 children)

Even if it was, there’s no way to know, people can just lie. It’s not like it will be obvious, some people might have a feeling it is (based on their experience playing with LLMs) but won’t be able to point exactly why.

[–] balder1993@programming.dev 2 points 2 months ago* (last edited 2 months ago)

...so might as well say that "agent" is simply the next buzzword, since people aren't so excited with the concept of artificial intelligence any more

This is exactly the reason for the emphasis on it.

The reality is that the LLMs are impressive and nice to play with. But investors want to know where the big money will come from, and for companies, LLMs aren’t that useful in their current state, I think one of the biggest use for them is extracting information from documents with lots of text.

So “agents” are supposed to be LLMs executing actions instead of just outputting text (such as calling APIs). Which doesn’t seem like the best idea considering they’re not great at all at making decisions—despite these companies try to paint them as capable of such.

[–] balder1993@programming.dev 4 points 2 months ago* (last edited 2 months ago)

If they actually wanted quality documents for people to use, they would be advocating for Standard Ebooks or something.

Or… you know… have PDFs that aren’t pictures of handwritten text?

[–] balder1993@programming.dev 2 points 2 months ago

Yeah, none of that would be a problem if the car isn’t connected to anything (WiFi, Bluetooth etc.)

[–] balder1993@programming.dev 2 points 2 months ago* (last edited 2 months ago) (1 children)

Imagine when the author hears about the “Create an app in 20 minutes with AI” tools.

[–] balder1993@programming.dev 4 points 2 months ago (2 children)

This is why technically software is a liability. The less code you need, the better, since every line of code is a potential vulnerability and something to maintain, update, etc.

[–] balder1993@programming.dev 1 points 2 months ago

In a few months, macOS will not even support Intel.

That’s a bold statement.

[–] balder1993@programming.dev 3 points 2 months ago* (last edited 2 months ago) (1 children)

But is it that different than the podcasts voices Google already generate with NotebookLM since a while ago?

[–] balder1993@programming.dev 4 points 2 months ago (1 children)

Only chatGPT has these kinds of comments as if you’re seeing code for the first time. 😆

I’m not against adding comments where is needed: in the company I work for (a big bank) my team takes care of a few modules and we added comments on one class that is responsible to make some very custom UI component with lots of calculations and low level manipulations. It’s basically a team of seniors and no one was against that monster having comments to explain what it was doing in case we had to go back and change something.

For 99% of the code you just need to have good names though.

[–] balder1993@programming.dev 1 points 2 months ago* (last edited 2 months ago)

Just a reminder that reencoding already compressed videos is a recipe for destroying the quality, unless you’re using a very high bitrate, which quite often gets you the same size as the input video.

I think the consensus is that if your video isn’t 4k or higher, there isn’t much gain in using HEVC if it is already H.264.

So if you want to store them long term, reencoding them now means that if you decide to do it again later (for whatever reason) you’ll have too many artifacts accumulated.

[–] balder1993@programming.dev 1 points 2 months ago

Yeah, a video is a video.

view more: ‹ prev next ›