I’ve been a long time Redditor and an Apollo user for about a year. I even paid for it. The main draw for me was the lack of advertising. In the back of my head I kept thinking that it couldn’t last. Reddit is losing revenue from the lack of advertising views. It didn’t
To me, Reddit’s sky high pricing for the use of the API is intended to kill off apps like Apollo and for its users to move to the advertising filled web site or its own app, which I’ve never used.
If Huffman came out and said this was a revenue move right off would everyone be as upset as they are? Are people upset because Huffman completely mishandled the move or because they got their ad free experience turned off? If Reddit had an app the same quality as Apollo only with ads, would they be OK with it. I’ve only used Apollo so I can’t speak to the other apps.
I can’t blame Reddit for wanting to make money. It doesn’t make a profit. Investors have to keep pouring in money to keep it going. They’re going to want to see a return on their investment at some point. Usually they cash in on an IPO, but IPO’s are generally only successful if the corporation looks like it will be profitable or at least the stock price continues to go up. That’s how capitalism works.
In my case, I probably would have left regardless. I can’t stand adds in my feed. I probably wouldn’t have heard of lemmy or kbin if there hadn’t been such an uproar. So I’m glad it went the way it did.
I don’t disagree with your saying that is why they’re doing it - but I wonder how well it will actually work out for them. Natural Language Prompting is hard to “lock in” someone on. Sure, the complex jobs with custom trained models are going to get locked in for sure - but the companies that are just adding “chat bots” to their apps? I don’t see the difficulty in migration.
I use OpenAI for one of my projects, and frankly there’s little that would keep me from being able to migrate to another service if one came along that gave a better value. An AI platform isn’t like an IaaS platform, there really isn’t a lot of platform-specific workflows involved, and prompts that work on one LLM should work just about as well on another.
Even for custom trained models, most training data is stored in json, and should be easy to feed into another LLM, though of course tweaking will be required