Would argue its about control too.
When you control the flow of information, you control how people will use that information.
Jokes on them, when there are no salaries there are no consumers, so no economy.
There’s an economy - paying rent to each other until there’s only one musical chair left.
Sounds like a commie agenda.
It almost is. You know they’ll run to the government demanding they give out stimulus checks, UBI and the like to the population to keep consumption going.
Some companies have already their own powerplants, then they can acquire some primary production and then finally they can be their own state and play King without subjects.
Which is funny, because even by that metric it doesn’t work since they’re hiring people to moderate the training data and output
A) expenses for that are pittens compared what may be saved in wages later on.
B) First gen of GenAI products, so development still very much ongoing. Of course more RnD is needed, which always costs money. Was the same with pretty much very product we can by today. Nobody in their right mind would expect the first shot to be the final product.
That being said, some ten twenty years from now though…
Don’t forget health insurance claims too
They pay people to find excuses to reject claims. Same problem, just applied to another problem.
It’s also the latest hype technology that consumers want little to nothing to do with.
After the big AI update on iPhone, I haven’t noticed much of a difference, and I’m honestly grateful for the limited extent to which it has been thrust upon me so far.
There are definitely legitimate use cases, but for the most part, this new way of generating text is confused with actual intelligence.
You’d be surprised. There’s definitely a ton of interest from consumers. Anecdotally, my wife used it to create a few quick logo ideas for her private practice (she hired a real artist for the final one), my coworkers and I use it for quick boilerplate script template creation, my immediate and extended family have all used it to modify/clean up family pictures, friends have used it in group chats for all sorts of things, etc etc. There’s a reason that it’s being implemented everywhere, and it’s not simply because there’s no consumer interest and it’s all corporate hype buzz. Just because you specifically aren’t interested doesn’t nullify the tens of millions of people using the various flavors of ChatGPT, Gemini, and/or whatever the hell Amazon’s is every day.
But yes, it is not real intelligence. I don’t think I’ve talked to anyone who truly believes it is. It’s just a new, highly versatile tool. Hell, I just saw a video of some jackass on YouTube programming a robotic arm to be controlled by ChatGPT and it had a rifle mounted to the arm. Using voice commands, ChatGPT was able to aim and shoot the rifle with crazy precision and speed: https://www.reddit.com/r/Damnthatsinteresting/s/popAFs2kmY
I’m not sure why you’re being downvoted stating facts, what is this reddit. There’s massive interest. https://fortune.com/2025/01/07/sam-altman-openai-chatgpt-pro-subscription-losing-money-tech/#:~:text=The company has projected that,power used to run ChatGPT.
There’s interest as long as they don’t charge the full value of the product.
What if it wasn’t $200/mo? What if people had to actually pay the cost?
Surely you understand no one likes working. It can do mundane, dehumanising and dangerous things.
AI can also solve other problems like premature death, illness, prevent crime, optimise food supplies etc.
This is much too optimistic for reality. Surely you realize it doesn’t matter what “one” likes or doesn’t like. It matters what the 1% likes. “Can solve” vs “will be allowed to solve”
AI could be used to automate menial tasks, but instead it’s being used to automate creativity. The people who own the AI also own the companies that benefit most from not paying people to maximize profits on creative work. They will not use it the way you and I envision it.