I see a lot of glorified weekend projects raising millions of dollars... pretty much anyone with little coding experience can put together a langchain + vector db backed app together so why would even small companies spend thousands of $ on a tool when they could easily build their own? Am I tripping lol would love to hear a different perspective
Ask HN: Is the AI tool market a too crowded now?
1–7 of 7 posts
Re: Ask HN: Is the AI tool market a too crowded now?
#2I think it's more about potential TAM than the current product. The hypothesis is probably that there will eventually be a winner in each vertical, and that $$ can mint the winner.
Re: Ask HN: Is the AI tool market a too crowded now?
#3For stuff that works “no”, for products where the business plan is “me too”, I think “yes”.
A key differentiator is making large evaluation data sets that prove that it works. Many hustlers are way too lazy to do that.
Re: Ask HN: Is the AI tool market a too crowded now?
#4I'm in the digital marketing space and I swear a new AI tool is being pushed on the daily. They are being ate up by so many people and also raising insane amounts of money. While I feel like 99% of the tools are absolute garbage, I believe that small business owners are feeling FOMO and wanting to stay involved.
Re: Ask HN: Is the AI tool market a too crowded now?
#5Can anyone name a nontrivial functioning llm tool, that actually helps in some business task? Not counting chatgpt, something using langchain + vectordb for example? I see lots of minimal examples but nothing real.
Re: Ask HN: Is the AI tool market a too crowded now?
#6I think it's more about potential TAM than the current product. The hypothesis is probably that there will eventually be a winner in each vertical, and that $$ can mint the winner.
Aah, that's insightful. I do believe there will be winners in some verticals but at the same time it seems like OpenAI is cannibalizing its own (SaaS) customers by making stuff like Code Interpreter available on ChatGPT itself...
Re: Ask HN: Is the AI tool market a too crowded now?
#7For stuff that works “no”, for products where the business plan is “me too”, I think “yes”. A key differentiator is making large evaluation data sets that prove that it works. Many hustlers are way too lazy to do that.
I agree!