What is wrong with people who pitch AI startups but they are really AI wrappers?

One buzz word that I keep reading here is the "AI Wrapper Startups". What are some examples of these that you find interesting? Please add 1-2 lines to describe the use case.

Seeing this for a few weeks now. Noticed it when PG tweeted about how there are 100 ai startups in the batch and all are legit. That is none of then it’s phony or opportunistic. He also talked about how domain experts are building in AI and great founders. Then saw Garry Tan promoting the batch and retweeting others doing it. A couple of other partners talked about the companies they are excited about too. To me, this signals that yc isn’t getting the kind of interest they used to get and they have to promote the quality of the batch to be able to generate enough interest. For the first time in 10-15 years maybe. If they have to do it to ensure the companies are getting enough funding interest, what about startups without a brand name behind them. We will see more of these as we get close to the demo day. Most startups are just ai wrappers or agents. Not sure about the exact differentiation between them. Not clear from the descriptions or websites.

What would happen to all the AI startups that are raising massive rounds at insane valuations once OpenAI improves GPT-4 (some spicy rumours are going around) or comes up with GPT-5? Most of these "AI" startups are just a wrapper around these existing, capable LLMs. Better versions of these LLMs would start generalising more and doing those specific functions done by these startups much better.

What is the reason for this? Lack of pay, Lack of cutting edge research?

I've been curious about whether there are any good Indian startups in the AI field that are not building "tools" like meeting summary or anything of that sort. But more like research based such as OpenAI, Anthropic, Midjourney, etc. They could be early stage or late stage, doesn't matter. Have you guys seen any around you with potential?

Was looking for automation solutions for my startup, came across an early stage AI startup that had 3-4 meetings with us and showed us screenshots of their product, but delayed product demo and signup. Few days later said they're shutting down because they couldn't raise funds and had no MVP ready to use. But they didn't tell us that earlier, all they said was "we are in early stages, still working on product, need info from you"etc. While their vision seemed good enough, not having an MVP while pitching for clients is weird and a waste of time tbh. No one will know or learn anything unless they try using an actual product, including themselves. And it isn't hard to build one either, i think I have the gist of it and might already be in production by other startups in this space. Decent low hanging fruit in SaaS with good long term potential. Doing things that don't scale doesn't mean you don't put in the required amount of work before approaching clients. Don't want to name them, they probably learned as much by now.

Something I've found very frustrating while pitching to investors, platforms and funds is that so many of them actually just bait and switch. They pose like gyani people earnestly looking for investments or helping entrepreneurs, but when you speak to them they start selling you their services. Fundraise is apparently just a step away, need to signup here first, or create xyz report first, or do outreach to xyz number of people, or do market research and fresh pitch deck creation, or hire xyz consultant/agency to get to the "next level" - basically a bunch of barriers you need to pay to clear despite having a profitable business with paying customers. Fact of the matter is there are plenty of people raising money based off of pitch decks with no MVP , or even just the right conversations with the right people. But they are rare and hard to find organically. Requires lots of networking and connecting dots over months and years. The point is, this selling in the guise of securing investment is tiresome and a waste of time for most entrepreneurs. They probably do get a few suckers occasionally which helps them stay afloat I guess. Some of them might even take your ideas and start building something of their own or back someone to build it under them. Once you start pitching, the deck starts floating around in VC groups and networks mostly beyond the reach of the average entrepreneur. Considering all this, just keeping one's head down and bootstrapping forever in peace seems like a better approach.

•⁠ ⁠One thing is clear - AI is going to take away jobs, and not just engineers. •⁠ ⁠So many "AI & SaaS" startups are just plain OpenAI wrappers. Their teams know what can be done "faster, better,smoother" but don't know how. One company even credited them in their poster. Well, honest at least 💁‍♀️ •⁠ ⁠They let in too many nibbi/nibbas and there were queues everywhere •⁠ ⁠People selling skin brightening creams in 2024 is still very weird to me •⁠ There's an electric cycle startup called Booty (i'm not making this up 💁‍♀️) •  In all, the management was very good. Food was expensive though 😢 •⁠ ⁠Best is the picture below, how Biryani tends to find me. This time the dogs are having fun too!

- Sarvam AI - $41M (Lightspeed and Peak XV) - Mad Street Den - $67M (Avatar, Peak XV) - Wysa - $25M - Neysa Networks - $20M (Matrix Partners) - Scribble Data - $2.3M (Blume Ventures) - Expertia AI - $1.3M (Chiratae Ventures) - OnFinance - $ 1.1M (Silverneedle Ventures) - Krutrom AI - $50M (Matrix Partners)