Ask HN: Is anyone else bearish on OpenAI?

The underlying technology of LLM's and stable diffusion is an interesting topic that does have some useful applications and should be studied more, I just don't see this company (or any others for that matter) GPTing their way to AGI within our lifetime, or being able to create significant value for investors after the hype fades.

This feels a lot like crypto where everyone is very excited about a new technology that very few people really understand, and are jumping on the bandwagon without asking any questions.

It's also very much like crypto where for every one person doing something useful with it, there are 20 trying to exploit the newness and low comprehension the general public have of the tech such as:

    - Trying to cash out on a ChatGPT wrapper company

    - Creating the nth "AI powered custom chat bot but for x vertical"

    - Using it to cheat on school assignments or interviews

    - Gluing together as many different "AI" services as possible to create a no touch business and sell low effort products

I'm not saying the company will go bankrupt but I'm also not buying into the hype that it's going to become the next Google or better / create AGI for us all.

What am I missing here?

AGI is a big ask.

Revolutionizing how we interact w/ computers by allowing us to use plain human language to do things the requester does not understand how to do seems to have been demonstrated. See even the relatively simple agent demo where an architect used human language to have zapier take action based on meeting conflicts. imo this alone is a big deal.

LLMs are a type of intelligence, and that is super useful.

Even without AGI, there are lots of expensive to scale currently tasks like customer service that it could take over.

OpenAI, at least in my day-day workflow for the last 9+ months has so superseded anything that google ever was to me that I'm having a difficult time comparing the two.

I've got a monitor dedicated 100% of the time to ChatGPT, and I interact with it non stop during the flow of technical scenarios and troubleshooting situations that flow into me - working in areas that I have the slimmest of backgrounds in, and shutting down, root causing, and remediating issues that have been blocking others.

I've essentially got 15-20 high-priced world-class consultants in every field that I chose to pull from, working at my beck and call, for $20 a month? I would pay $200/month in a heartbeat out of my own pocket, and I probably would ask the company to pay ~$2,000/month for my workflow.

I think if they never released another product, and they just managed to penetrate with their existing offering, they are easily a $100B+ company once they nail down how to monetize.

The difference between LLMs and Crypto is I can point to roughly 200-300 objective solutions over the last 9 months where ChatGPT resolved an issue and delivered clear value for me alone. And, over time, as you learn how to control for hallucinations, and manage your query patterns a bit more - the value has continued to increase.

That same multiple-times-a-day high value persistent experiences were never a part of my crypto experience.