Meta planning sweeping layoffs as AI costs mount

AI is super energy intensive, as is manufacturing the chips necessary to run it. An energy crisis will make it even more of a money loser than it already is.

We got a guy at work who graduated last year and did really well in his degree. But he can't write well and doesn't even seem to know basic things about his subject. He admitted to using his AI for his assignments, but I have no idea how he was able to pass his exams. My sister's at uni and she's also mentioned that everyone is using AI for their essays. I agree, I don't think students at least will stop supporting AI. It's a shame because it defeats the point of getting an education. 

I see the danger but I’m able to be much more productive at my job while taking away most of the frustrating parts

its absolutely going to be damaging to the workforce for the foreseeable future. but yeah, the insane costs they're throwing to make it sustainable is digging them deeper and deeper, and outside of higher profits in the short term (because they're paying less employees), there isnt going to be a long term benefit. consumers already realize AI is c*** for the most part. and most of us are learning to tighten our wallets for fear of being replaced, so *the people* arent going to buy in for multiple reasons, while corps continue to go deeper and deeper into debts hoping for something that can clearly never happen.

Your company is using LLMs for *accounting?*! What's the name of the company, so I can make sure I never use it. Jesus.

Circle of life, the company who will replace AI will either have a shittier product and slowly be overran by competitors, or hire back the employees. Of course there are roles that could be replaced with AI, but in my experience this type of jobs is already outsourced to poorer countries.

AI is making certain roles obsolete for sure and that will continue to happen. As for replacing entire departments, there’s probably going to be a lot of regret. At the end of the day you can’t hold AI accountable for anything, AI isn’t intelligent like a human is (not that CEOs can tell the difference), AI derived work can’t be copyrighted, and you’ll have the doorman fallacy in full effect. Like remember when Martha was laid off last year and now you’re essentially doing two jobs, well management probably didn’t realize that when they laid you off. While some companies will adopt AI in a beneficial way, many are just setting themselves up for lower sales and long term increased costs. Like with AI you can automate an entire sales team, but with every company doing that and spamming the living f*** out of everyone, that’s not going to increase sales. AI slop is going to kill digital sales. People will ignore their inboxes (more than they already do). Funny, but it might be the companies that still employ humans that benefit the most since they are the ones that will stand out. It’s humans that can wine and dine you or meet you in person. AI isn’t going to go around to clients making sure everyone is happy and ready to renew. Personally, AI is setting up an environment for the next wave of disruption. Companies built by small teams, utilizing high quality AI to level the playing field with massive corporations offering the same services at insanely low prices. Every time one of these companies is bought out, two rise in their place. AI could be the monopoly killer we’ve been waiting for. Why overspend on enterprise level software and services when you can build your own?

It's a force multiplier for people who already know how to do their job well. This will make things difficult for junior positions because there will be less demand for them, and they might not have the experience to notice when their AI goes off the rails.

>If you can't filter stuff you read online based on your own knowledge or verify/test it out then a normal search is also useless. It is not. Normal search is very useful to find information that you don't know about. An easy case I tried a few days ago. I needed to find the different rules followed by my regional and national government to obtain a certificate, to evaluate which way more convenient. I tried both regular search (with no AI capabilities) and GenAI. Regular search just took me to the government pages with the documents describing their procedures. They are very annoying documents to read, but the information is there and it's trustworthy. I don't need to know about the regulation beforehand to be able to trust it. I then tried AI, just to see if it could give me a distilled version highlighting the important points. It did, but the information was wrong in ways you wouldn't know unless you compared it with the actual documents. It mixed different certificates and agencies in the same result. It made up documents. It wasn't just the difference that was wrong, but the actual requirements from both agencies were wrong. I noticed it because I was doing the regular search side by side, but it looked completely believable. Once I pointed everything that was wrong, it eventually told me it couldn't access the specific documents in which the regulation was described. Someone treating the gen AI output as a regular search result, and trusting it the same way they would trust a regular search result, would face some nasty surprises. The first thing all these algorithms should do to be able to use them as search engines would be to NEVER hallucinate, and prioritize no results over wrong results.

I never said banning AI. I said ban AI practices, via laws, to protect art privacy, and some forms of work. At least as temporary measures until we have a chance to adjust.

This is used for fertilizer and oil prices at his company

Hey guys, I found him. It's this guy's fault!