Ask HN: How many AI startups are just OpenAI/Anthropic/etc. API calls with a UI?
I'm curious about the current AI startup landscape and how many companies are essentially just wrapping foundation model APIs (OpenAI, Anthropic, etc.) with a UI layer versus doing more substantial technical work.
Some questions I'm interested in: - How can you tell if a company is primarily using foundation model APIs? - What percentage of AI startups fall into this category? - Are there examples of companies doing this particularly well or poorly? - What constitutes legitimate value-add on top of foundation models?
I spent a good chunk of the pandemic doing startup advisory, and some of that has lasted up until the holiday season, so I can tell you that there was a marked shift one or two years ago where the investment board I had a seat on decided to point blank stop looking at startups that were solely focused on AI, because even the non-technical board members realized that there was little to no added value there -- and we were getting like 70% of new prospects in that category.
I would go into a demo, look at it, ask them how they did RAG on the data, ask to speak to the people doing their AI models, etc. And then sometimes I would spend a couple of hours wiring up a Node-RED flow to show my colleagues how trivial it was.
The stuff we (now they, since I'm taking some time "off" that moonlighting gig to recover from burnout) ended up prioritizing are companies that focus on business processes where triage and "human augmentation" can actually benefit from LLM summarization, some automated decision making, and some data "integration" (not just summarization, but broad correlation of events, etc.)
There's a _lot_ to be done in many fields where, say, you will notice a spike in some piece of data (using conventional ML), gather the data around that event and present it to a human (with pros and cons, including trying to flag if the data is reliable). Think GitHub issues for crop management, and you're halfway there.
_Those_ companies really need to have their use cases sorted out, and not just try to be "the Uber for greenhouse management" because they wrapped weather forecasts into an LLM.
So, in short, the real added value is expediting or improving use of domain expertise captured from (or still held by) humans.
> So, in short, the real added value is expediting or improving use of domain expertise captured from (or still held by) humans.
I agree here, AI-infused automation/orchestration of sometimes long-running business processes with a human in the loop seems to be very high value.
> decided to point blank stop looking at startups that were solely focused on AI
The latest ycombinator AI promo video on youtube makes it seem like few are getting in anymore unless they are building around AI. And they strongly push that startups should be hiring developers who use AI tools on top of that.
Well, I'm in Europe. We tend to have different assessments of value.
I think as models get commoditized, that doesn't really matter. In fact, the UI is what becomes important. We can already see this with Deepseek and other Open-Source models.
I've heard this idea echoed quite a few times.
But can anybody name a wrapper company as/more valuable than foundation model companies like OpenAI, DeepSeek, Anthropic, Mistral, etc?
Only one that comes to mind is Perplexity, but they're a bit more than a wrapper startup - I think there's some hardcore engineering to get their web search product working so well.
Of course there aren't any that are more valuable.
But take Notion, Asana, Trello. All of them started out.. like a UI around a database. And they're plenty valuable now. More valuable than Oracle, the highest valued DB company? Maybe not, but still very much successful companies. They didn't start out with 500 integrations.
I think Character.ai is already very profitable, and that's a pure wrapper.
I do wonder after deepseek bombshell how can any one come up with an idea, start building, and not be constantly second guessing themselves on "has the industry changed this week making everything we are doing moot?"
Its almost like you need a full time AI researcher on team, not to help build, but to keep up on all new info and constantly tell the team PIVOT!
Why does this matter? MS CoPilot uses OpenAI. One view is that if a foundational model is your moat, you have no moat. Even using commodity AI, you can still lead providing a service built on it.
I mean, how many SaaS startups are just UIs around Postgres databases?
How many startups are UIs around rsync?
What are some truly valuable companies that would be accurately described as just "UIs around Postgres"?
I think it's a non-sequitur - either companies are much more than just UIs around a DB, or those companies just aren't very valuable (in which case, who cares).
The exact same holds for the pure AI wrappers - they'll end up similarly valued to UIs around Postgres.
Only difference is those startups don't rub it into your face they're database-based.