Phind is better about linking sources. I’ve found that generated code sometimes points me in the right direction, but other times it leads me down a rabbit hole of obsolete syntax or other problems.

Ironically, if you already are familiar with the code then you can easily tell where the LLM went wrong and adapt their generated code.

But I don’t use it much because its almost more trouble than its worth.

There are options for additional checks they could explore that are less creepy.

1. Make someone pay a one time charge. This might not give you the full details but the transaction gives more data, and shows someone is willing to back up their request for access with a paper trail which has fraud protection laws.
2. Third party verification services. Like your bank, who already have details about you. There just needs to be a way for them to vouch for you. Credit reporting agencies probably already do this, but I kinda think this is almost creepier than giving Facebook a video.
3. Verifying the email attached to your account is a good first step.

In the meantime I think knowing the password should at least get you logged in enough for account maintenance. You should be able to set the entire account private and take it offline with limited toggles. Restoring full access would require the additional verification.

They have tasers, which didn’t work. Then move to pepper spray to blind the person and batons to subdue them.

Other countries where guns are uncommon but knife attacks are more common they have long hooked poles to capture someone. That’s impractical to carry so wouldn’t have been available in this situation but it is a tool that can be used when called out to a knife threat.

The problem here is the response doesn’t match the crime (fare avoidance) and doesn’t match the threat either. If the only solution you have for littering is to blow up a city then its time to step back and rethink the problem.

Three innocent people injured, months of investigation, and millions in lawsuit settlements. This is a failure of the officers but also a failure in training them how to respond.

Just want to point out that it absolutely is possible to train an AI that will keep track of its sources for inspiration and can attribute those when it makes a response.

Meaning creators could be compensated for their parts of AI generated stuff, if anyone wanted to.

The free software as a passion project idea became untenable long ago. It works for UNIX style utilities where the project stays small and changes can be managed by one person but breaks down on large projects.

As a user, try to get a feature added or bugfix merged. Its a weeks or sometimes months/years long back and forth trying to get the bikeshedding correct.

As a maintainer, spend time reading and responding to bug reports which are all unrelated to the project. Deal with a few pull requests that don’t quite fit the project, but might with more polish. Take a month off and wait for the inevitable “is this being maintained?” Issues reports.

I contribute back changes because I want those features but don’t want to maintain a longterm fork of the project. When they’re rejected or ignored its demoralizing. I can tell myself “This is the way of open source” but sometimes I just search for another project that better fits my needs rather than trying to work on the one I submitted changes to.

That is the happy path. The sad path of this is how many people look at the aforementioned problems and never bother to submit a pull request because it’s too much trouble? Git removed most of the technical friction of contributing, but there is still huge social friction.

Long story short: the man pages maintainer deserves something for all the “work” part of maintaining. He can continue to not be paid for the passion part.

If we keep going we might accidentally reinvent Usenet news.

Not saying that like its a bad thing, just saying we might be able to take some inspiration from there.

Can do… But won’t

https://m.youtube.com/watch?v=rsDeBrtY5W4