3 ms·
I believe the intoxicating idea that companies will not need employees soon is driving more of this than we think. Once the hype wears off things may normalize
by osigurdson 4y ago
I believe the intoxicating idea that companies will not need employees soon is driving more of this than we think. Once the hype wears off things may normalize again.
I say this as a heavy chat GPT user. I really love it but not nearly like sending and email to a senior engineer: “please make X” - which then turns into the senior engineer scheduling meetings with the required teams / individuals - navigating the technical and (unfortunate) political issues, coming up with a design and implementing it - getting things done in other words. I don’t know what future versions of GPT will look like but the current one is light years away from having that level of capability.
Even a simpler task like fixing a bug cannot remotely be handled by ChatGPT.
It is almost like LLMs are a new dimension - we used to have X and Y, now we also have Z. Yet, we are thinking that this is just more “Y”.
- whatyesaid 4y agoIf it was about LLMs then it wouldn't have to be preemptively rushed. Like, they could wait a year or two, their bank account isn't going anywhere. The users and data of big tech are the real value. By waiting a year for LLMs to take effect you evaluate properly, long term lose nothing, and don't have to go through massive hiring to compensate. It's not about LLMs. The real reason is these companies have been overpopulated and produced basically no visible changes in 5 years. Google search is worse with auto-generated results not written by humans, YouTube has become super rabbit hole instead of discovery, Twitter shipped NFT profile pictures and Spaces.
- osigurdson 4y agoI think the Occam's razor hypothesis is the stated one: economy looking cloudy, money getting more expensive. However, I'm certain LLMs are part of it as well - when you are the one writing the cheques you certainly do want to believe in a future where no more cheques need to be written while maintaining the same revenue. I suspect it is more likely that the benefits will accrue to start ups as they will not have so much inertia working against them in terms of leveraging AI. People with ideas and limited means will likely be the primary benefactors due to AI amplification I think.
- mulmboy 4y ago> Twitter shipped NFT profile pictures and Spaces. Don't forget fleets rip
- eli_gottlieb 4y ago> I believe the intoxicating idea that companies will not need employees soon Look, if you want to not need employees, you need to do capex. If you spend more money on stock buybacks than you save in payroll by doing layoffs, you're taking money away from capex.