AI & Gen AI News Watch: 2/23/24

Here's the top gen AI news this past week:

• This appears to be a week for weird AI. As mentioned above, Google pulled Gemini for image issues, and earlier in the week, OpenAI basically had to "reset" ChatGPT/underlying models as they started spouting nonsense. 
• On a positive note, Open AI announced that it's "improved the memory" of ChatGPT chats, basically making ChatGPT more intuitive and user-friendly. 
• On Wednesday, Google announced Gemma, an open source version of its Gemini LLM (formerly known as Bard). (BTW: If you’ve ever wanted to play with an LLM directly and you're not a data scientist, there's a nice guide for getting your hands dirty with Gemini available here). 
• Looking for examples of how companies are using AI? How about this one exploring how it’s being used to accelerate mining discoveries. 
• Or this one creating an AI biology model. 
• Or this one discovering new materials. 
• Or this one about a small weather prediction company bypassing competitors thanks to its use of AI.
• If you're looking for a new way to use generative AI in your company day-to-day, Adobe just added the ability to "chat" with documents to its PDF tools. 
• Of course, adding AI tools isn't always a slam dunk, as JetBrains recently found out. They recently removed an AI assistant from their programs (and are being praised for doing so). 
• Speaking of AI technology possibly gone too far, employers seem to be gravitating to (and candidates seem to be pulling away from) these new AI-driven personality tests fused into online job applications (although, to be honest, we're not sure we see what the AI angle is here). 

Posted by Becky Nagel on February 23, 2024