Sam, just add sponsored content. The road to enshittification doesn't have to be long! Make it shitty fast so people can move past it and start hosting their own models for their own usage.
$200 a month for a user is losing money? There's no way he's just including model queries. An entire a6000 server is around $800 / month and you can fit a hell of lot more than 4 peoples worth of queries. He has to include training and or R&D.
Much like uber and netflix, all of these ai chatbots that are available for free right now will become expensive, slow, and dumb once the investor money runs out and these companies have to figure out a business model. We're in the golden age of LLMs right now, all we can do is enjoy the free service while it lasts and try not to make it too much a part of our workflow, because inevitably it will be cut off. Unless you're one of those people with a self-hosted LLM I guess.
You can help by asking ChatGPT to produce the most processor intensive prompt it can come up with and then having it execute it repeatedly. With the free version this will burn through your allotment pretty quickly, but if thousands of people start doing it on a regular basis? It'll cost OpenAI a lot of money.
So people are really believing Altman would publish these damning statements without ulterior motives? Are we seriously this gullible? Holy shit, we reached a critical mass of acephalous humans, no turning back now.
This 100% answers my question from another thread. These businesses have cooked the books so bad already that they thought this was gonna save them and it doubled down on em.
sam altman proving once again that he is not only a tech genius but also a business genius. make sure to let him scan your eyeballs before it’s too late.
What are people using the $200 plan for that makes it worth it? You only get their model with their training, you don't have any access to weights or training. And with how nerfed openai makes its models, nothing even remotely nefarious can be done with it. All you can do is process simple data. Which having a purposed trained model seems the most valuable for.
I cancelled a 20$ subscription I started because it was arguably useful for me and served exactly one use-case. Now I don't need it anymore.
Of course, they had a form asking feedback/why. I chose "ChatGapT is nott advanced enough" as that was one of the alternatives. Hopefully it will lead to them putting more resources into development and burn through investor money faster.
If they are losing money on $200/month, that does not necessarily mean they lose money on the $20/month.
One is unlimited, the other is not. You only have to use the $200 subscription more than 10 times the amount the $20 subscription allows for OpenAI to earn less on that subscription.
Hmm, we should get together some funds to buy a single unlimited subscription, and then let it continuously generate as large and complex prompts as the rate limitting allows.
This is something I've been speculating for a while. The cost of running these complex systems (as OpenAI models aren't just LLMs) is subsidized so heavily that we don't really know the cost of running these things.
This is a huge risk to any business, as the price for these services has to go up significantly in the long term.
Wonder what their financials are actually, usually with cloud capital ventures like these they usually still magically keep raking it in even though they're "losing money".
For instance, Amazon during the pandemic paid zero corporate tax even though they had record sales because they "didn't make any profit", Tesla too who didn't make profit during 2020 yet their share price went 10x and they had plenty of shares to sell if they wanted.
This reminds me of this fossil cable company owner. He just couldn't understand why they needed more bandwidth. Netflix and youtube had come in to being and this old fart just didn't get that people were actually using their bandwidth.
Okay everyone should create an openai account and start feeding it shit. Ask it the most braindead questions ever, if they use your questions as training data itll just fuck the next model up even more.
Yeah, we see this all the time with emerging tech and platforms. All of the top tech companies now were once spending cash faster than they could make it, and all the naysayers saying they’d never be profitable.
Don’t y’all get tired of being wrong sometimes? Maybe try to learn from the past.