OpenAI has been pretty damn good at closing their weaknesses. Originally AI was hard to communicate with, they released instruct models. It couldn't hold conte

OpenAI has been pretty damn good at closing their weaknesses. Originally AI was hard to communicate with, they released instruct models. It couldn't hold context and deal iteratively, so they released chat models e.g. ChatGPT. AI can't do math? They added code interpreter so it would code its own calculators. It couldn't access APIs, they added functions. Companies like Voiceflow would specialize in building assistant bots, so OpenAI released assistant. They're able to build fast, like a startup, but they're funded far better than a startup. So it's going to be hard to close in on these opportunities. But if I have to answer the question: 1. Language and dialects - Indonesia has about 700 languages, most with significant overlap, but it would be nice to have something that supports more of them. 2. Safety. It goes overboard with it, which I think won't work anyway in the long run. A bot that could swear is nice but I don't think they'll go here anytime soon.