Hacker News Reader

Comment by m_ke | Hacker News Reader

I wonder how much this shift to enterprise software is due to YC having a large portfolio of enterprises now. It's a lot easier to build a B2B business when you can get an intro to all of the unicorns that came out of YC.

A company like Scale AI can get Airbnb and OpenAI to play with their service, then throw their logos on a deck and go sell it to Liberty Mutual.

View on HN

Scale AI's secret sauce is in labeling lidar point clouds, which is really a necessity for the self-driving car industry. Sure, they can label other stuff as well, but it's not hard and there are other companies that can do it (not quite as well, but they could easily catch up).

What I'm trying to say is Scale AI is pretty much the only company offering a solution for an absolutely hair-on-fire problem for self-driving car companies - the introduction was unnecessary (and many of their customers are Argo, Waymo, Zoox, and other companies not affiliated with Y Combinator). Assuming the self-driving bubble deflates (it will) then Scale AI will suddenly find itself a commodity business as labeling images is not so hard that it can't be easily copied. It will be a race to the bottom. Unless there is another sudden surge of demand for labeling lidar point clouds.

Yeah I know what they do and know a few other people working on data annotation tools for autonomous vehicles. All of the data annotation startups pivoted to that space because of the money involved.

I just used them as an example because they were the first recent YC company that came to mind that's working on enterprise tools and has a landing page full of logos.

Fair point. Sorry if I came across as argumentative - I just thought it wasn't the best example.

Regarding your overall point of YC having a network effect - not only is it true, it's part of their thesis for scaling up YC as much as possible. YC made a big push when Sam Altman became president (he's now heading up OpenAI) to try to grow YC as much as possible. This was about 5 years ago. The real question is how will the network effects weather the storm of a recession, which is untested (YC really started right around the last recession). Are they inside their own echo chamber? Are they actually supporting self-sustaining businesses? Will the network become big enough that it can sustain many of its businesses?

I'm curious to see how the tech startup scene shakes out with the next recession. Let's just say I'm not considered an optimist.

Does Scale AI use trained human annotators to label point clouds for customers or do they have a trained model that they use to quickly and automatically label point clouds for customers?

Using an ML model to label point clouds for generic ML purposes sounds like a horrible idea.

I met an ex-employee, they have 500 employees in the Philippines doing annotation. They built some software to help annotate point clouds but it seems like mostly a contractor handling logistics that the tech companies would rather not deal with, and they became a unicorn because AI.

Oh okay, thanks. I wasn't sure if they were selling annotations that their trained model can do quickly or selling manual labeling services that their customers can use to train their own models. Sounds like the latter which isn't as technically impressive.

I also had a similar experience with a YC company providing hair loss solution in India (nonucare.com)

On the website they have testimonials from founders and CEOs who vouch for their product. 5/6 of them were founders from their YC batch.

At some point you have to wonder if the reviews are genuine or just a case of founders helping founder friends.

>At some point you have to wonder if the reviews are genuine or just a case of founders helping founder friends.

It's a hair loss product. Of course they are not genuine.

Is that really a YC company?

> Used by 90% of USA

Uh huh.

I bet this effect is intended. The YC ecosystem can help itself grow and allow nascent companies to iterate more quickly by getting tech-savvy customers, feedback, and cash flow earlier. This can grease the wheels of the machine on the path of building a business from a conventional bottom-up approach accomplished by supporting SMBs and growing to support enterprises after finding product-market fit and iterating on the product and business rather than going directly to enterprises (where the sales cycles can be very long).

Companies evolve toward what they (think they) are good at. Probably something to do with min-maxing.

But if you shift left, right, up or down enough you leave a power vacuum, and your observation has me wondering if someone will step in to fill the space YC is 'vacating' (less focused on)

Reply on news.ycombinator.com