GPT-4o Discussion Thread

OpenAI have released a new model. It’s claimed to be around as intelligent as GPT-4 but much faster. It can transcribe in real time, and respond to a human conversation in a natural sounding way complete with emotions. It can see the world via your camera, hear it via your microphone, and does a pretty reasonable job at it.

Their YouTube channel has a bunch of live demos


And if you’re lucky, you’ll already have access to the model. Here’s a recording of their announcement event ft. live demonstrations /​ roll out schedule.

Discussion Prompts

1. How much progress has been made relative to expectations and how does this affect which timelines are most likely?
2. How can we use the new capabilities to save time /​ money /​ lives?
3. What industries is this going to disrupt?