I am wonder why leftists are in general hostile towards AI. I am not saying this is wrong or right, I just would like someone to list/summarize the reasons.
I am wonder why leftists are in general hostile towards AI. I am not saying this is wrong or right, I just would like someone to list/summarize the reasons.
Yes, I’m left-leaning, and I dislike what’s currently called “ai” for a lot of the left-leaning (rational) reasons already listed. But I’m a programmer by trade, and the real reason I hate it is that it’s bullshit and a huge scam vehicle. It makes NFTs look like a carnival game. This is the most insane bubble I’ve seen in my 48 years on the planet. It’s worse than the subprime mortgage, “dot bomb”, and crypto scams combined.
It is, at best, a quasi-useful tool for writing code (though the time it has saved me is mostly offset by the time it’s been wrong and fucked up what I was doing). And this scam will eventually (probably soon) collapse and destroy our economy, and all the normies will be like “how could anybody have known!?” I can see the train coming, and CEOs, politicians, average people, and the entire press insist on partying on the tracks.
when copilot came out and it was nothing more than an extremely fancy auto complete.
that was peak, I’d still write the logic and algorithms and the important bits, it just saved time by quickly writing the line when it got it right, it all went downhill from there.
I prefer using LMMs for tech debt stuff like starting a readme and doing comments.
I do the real brain work and the end product looks nicer.
Sudo code (baseline comments), real code, dev/test, LMM to add more words after. Smack it when it touches my code.
LLMs can be a great assistant. it’s like having an intern doing the tedious work while you get to just approve and manage it.
but letting it run the show is like letting the intern manage the whole development unsupervised.
Like the AI that was given access to prod, deleted a database and lied about it. The company also didn’t have a backup.
Source: Replit’s CEO apologizes after its AI agent wiped a company’s code base in a test run and lied about it
those are the fuckups that become public. there will be a lot of major fuck up.