4 ms·

I believe this is why ycombinator does not recommend building a business on top of another company's platform. I suspect OpenAI based businesses will eventually

I believe this is why ycombinator does not recommend building a business on top of another company's platform. I suspect OpenAI based businesses will eventually learn this lesson.

robertlagrant 4y ago
I imagine many would like to be bought ASAP.
illiarian 4y ago
It's fine to build a business on top of another company's platform. You do need a contingency plan, especially if what you offer can be easily replicated in that platform.
- realusername 4y ago
  There's no contingency plan possible on mobile unfortunately, you can't remotely switch customers to another platform.
  - mejutoco 4y ago
    A PWA could work in some cases, depending on the capabilities needed.
- gabereiser 4y ago
  Apple vs Epic Games is currently challenging this in courts over Fortnite. You can play on other platforms but the argument is on mobile, you're forced to use their walled garden and let them take 30%.
  - waboremo 4y ago
    Yes, and as per the very recent ruling, one would suggest that ycombinator is entirely correct in regards to not relying on a singular other company's platform, since they've decided that Apple is within their rights in regards to a walled garden.
    - toyg 4y ago
      The ruling still deemed illegal restrictions on making people aware that other options exist. But yes, the more one relies on someone else's platform, the more one is just a sharecropper.
  - scarface74 4y ago
    You’re a day out of date. Epic lost this case.
    - gabereiser 4y ago
      Appealed in appeals court. It's not over yet.
      - scarface74 4y ago
        It’s already lost in appeals court yesterday. The next step is the Supreme Court. Do you really think they are going to take it up?
- osigurdson 4y ago
  What would the contingency plan be in this case? It seems that it would just be to build something else compelling instead. In which case it seems that starting with plan B might be a good idea.
fakedang 4y ago
The number of YC startups this batch centered around OpenAI seems to belie this. Although I do remember YC stating this often in effect, at least in past, older content.
- osigurdson 4y ago
  To counter my own argument, the other possibility is OpenAI like features become a commodity. I.e. build a product around OpenAI and switch to Google's (eventual) offering at some point. However, embeddings might end up being the true vendor lock in of our time. If I've based my entire company on OpenAI embeddings (and spent lots of money on these), it isn't clear there would be an easy way to translate to another vendor's offerings.

I believe this is why ycombinator does not recommend building a business on top of another company's platform. I suspect OpenAI based businesses will eventually learn this lesson.

robertlagrant 4y ago

I imagine many would like to be bought ASAP.

illiarian 4y ago

It's fine to build a business on top of another company's platform. You do need a contingency plan, especially if what you offer can be easily replicated in that platform.

realusername 4y ago
There's no contingency plan possible on mobile unfortunately, you can't remotely switch customers to another platform.
- mejutoco 4y ago
  A PWA could work in some cases, depending on the capabilities needed.
gabereiser 4y ago
Apple vs Epic Games is currently challenging this in courts over Fortnite. You can play on other platforms but the argument is on mobile, you're forced to use their walled garden and let them take 30%.
- waboremo 4y ago
  Yes, and as per the very recent ruling, one would suggest that ycombinator is entirely correct in regards to not relying on a singular other company's platform, since they've decided that Apple is within their rights in regards to a walled garden.
  - toyg 4y ago
    The ruling still deemed illegal restrictions on making people aware that other options exist. But yes, the more one relies on someone else's platform, the more one is just a sharecropper.
- scarface74 4y ago
  You’re a day out of date. Epic lost this case.
  - gabereiser 4y ago
    Appealed in appeals court. It's not over yet.
    - scarface74 4y ago
      It’s already lost in appeals court yesterday. The next step is the Supreme Court. Do you really think they are going to take it up?
osigurdson 4y ago
What would the contingency plan be in this case? It seems that it would just be to build something else compelling instead. In which case it seems that starting with plan B might be a good idea.

fakedang 4y ago

The number of YC startups this batch centered around OpenAI seems to belie this. Although I do remember YC stating this often in effect, at least in past, older content.

osigurdson 4y ago
To counter my own argument, the other possibility is OpenAI like features become a commodity. I.e. build a product around OpenAI and switch to Google's (eventual) offering at some point. However, embeddings might end up being the true vendor lock in of our time. If I've based my entire company on OpenAI embeddings (and spent lots of money on these), it isn't clear there would be an easy way to translate to another vendor's offerings.