AI: Switch from OpenAI/ChatGPT to Claude/Anthropic or Mistral/LeChat

Title. Anthropic is the less evil of them all, with the best ethic and sycophancy of the big AI companies. It just makes sense to see an indie dev team collaborate with an AI company with aligned values.

There’s also Mistral with LeChat, which has the advantage to be based in the EU (France). And maybe we will see Proton adding an API to their own (recently released and privacy focused) LLM that is Lumo (that being said, the project is supposed to be fully open sourced - didn’t check the details yet on their GitHub).

In any case, Claude is much safer compared to ChatGPT. Claude is better for neurodivergent people. Is it better on everything? Nope. Still one of the new multi-billions AI company leading us to some uncertain future. But, at least, they’re trying to turn things around with an optimistic goal in mind. Which is not necessarily the case with OpenAI.

To most users, this might be a tad too technical, boring or unnecessary from their point of view. The thing is that, if you had the chance to compare ChatGPT with Claude, there’s a clear difference in the way things are phrased, structured, safer. I also have a strong feeling that a Claude is more suited for neurodivergent people. And I won’t mention the awful recent news of users relying on ChatGPT for therapy and stuff, leading some of them to extreme situation I won’t mention either.

Long story short: can Structured move away from OpenAI and either go for an open source AI backend (technically complex, I get that) or for a lesser evil? It can only improve the AI integration in the app. And offer better privacy control for every Structured users…. Even tho Anthropic is also doing some weird stuff - cf. their latest lawsuits and their 5 years data retention period for training purpose for individual users. It should be different on an API/Business level, but we all know that today, we can’t fully trust these companies privacy wise.