3 ms·
The Next Great Leap in AI Is Behind Schedule and Crazy Expensive
- jmathai 2y agohttps://archive.is/2024.12.21-093402/https://www.wsj.com/tech/ai/openai-gpt5-orion-delays-639e7693 https://archive.is/2024.12.21-093402/https://www.wsj.com/tec... I read a lot of these posts and get the impression that they are trying to make the point that AI isn’t transformational. The current suite of generative AI models is disrupting industries. And we’ve only had a couple years to build products on top of them. Sure there are limits. And in some cases we are already hitting them. But we haven’t had much time to think about ways to accommodate these limitations from either a technical or user experience perspective. We have a couple years worth of value extraction from today’s LLMs. Perhaps the investment into AI are banking on a timeline to the next breakthrough. But as it stands, even if we get not improvements from here on out we have only seen the start of how industries will be changed.
- fosterfriends 2y agoLink to non-paywall?
- jmathai 2y agohttps://archive.is/2024.12.21-093402/https://www.wsj.com/tech/ai/openai-gpt5-orion-delays-639e7693 https://archive.is/2024.12.21-093402/https://www.wsj.com/tec...
- Merik 2y agoThis is a hit piece. It is cherry picking information that is in the public domain, and then drawing spurious conclusions by purposefully excluding or dismissing other publicly available knowledge, and doing so with a tone of exasperated incredulity.