Recap: AI4 Conference Unveils Exciting AI Breakthroughs from Khan Academy and Google

0:00

Welcome to the prompt engineering podcast, where we teach you the art of writing effective prompts for AI systems like chat, GPT, mid journey, Dolly, and more. Here's your host, Greg Schwartz. Hello, everyone. I am back. So I appreciate the time that you were all able to give me to take off. It was very helpful. I'm on two new medications and they're definitely helping. Still working on dosage and a lot of other stuff, but it has made a huge difference. So I'm going to be publishing again. And this is the first one. So wanted to share. I just got back from AI four, which is an AI for enterprise conference and had a good time there, got to see a bunch of different people and see a lot of different talks. In fact saw a ton of different stuff. Lots of talks from people selling things to enterprise. A lot of non-technical talks as well, and a few technical ones. Unfortunately, there were Quite a few talks and products that were really a solution in search of a problem. You've probably seen that before. And frankly, you've probably seen this. Comic illustrating that before. And by the way people who are just listening. This is particularly an episode that you might want to go to YouTube to watch because I'm showing some different clips from the conference. But you'll still be able to follow along on audio. If you don't know, the YouTube channel is youtube.com/@PromptEngineeringPodcast. You can also go to PromptEngineeringPodcast.com to get the link to YouTube and to watch this episode. Did want to share. I saw a few particularly really good talks and a really interesting discussions. So the first one was A company called vital. Which is founded by the former founder of mint.com. They created a tool to translate. Doctor's notes and all the jargon and medical terminology into plain English. Now I interviewed them. So I'll save more discussion on that for later. But you're seeing at the moment. An example of me running some information through it and you can see it. Turns it into some actual normal English words. I also saw two talks on weather. Those actually had no chatbot stuff in the middle. They were like run of the mill machine learning and they were really cool. Two interesting uses that I had not heard of. One of them was able to predict wildfires. And also paths of hurricanes. And then there was another system Google's weather prediction system, which also was able to do. Hurricane prediction Next, there was a talk on bias. And you know how to deal with biases. In large language models. So examples of that would be things like. Biasing based on gender or race or religion or political ideology or profession and all these different things. And then They also, the speaker also talked about some of the ways that you can actually remove bias. Particularly that most ways of doing it are just, humans looking at it and going, hang on. That's biased. But people are starting to play with. Actually asking a model and an LLM two. Say is this biased, but there's some interesting technical challenges with that. And then there are examples of anthropic and other companies who are looking at. Creating what's called the constitutional AI. And I'm actually looking for a speaker to come on and talk more about that. So if that's something you know about, please let me know. And then my favorite talk was one given by the founder of the Khan academy. They've built a chat bot that doesn't simply give you the answer the way, you can ask chat GPT to solve problems or do your homework. Instead, it helps you to make better decisions and basically learn. The example that's being shown here is actually giving you feedback on how to write a better essay. But they also showed some examples of, for example, with a math problem, say. Y equals two. Times parentheses X plus seven close parentheses. The example shown was then incorrectly applying the times two to the pieces in the parentheses. And it then talked through how to actually do it right. And showed you the next step and said, could you explain how you got this answer? Because that's not what I got. Which I thought was an interesting nuance. It doesn't say that's wrong. It just says"that's not what I got, could you tell me how you got there?" One of the things that was really interesting about the chat bot was you could turn on a different mode. That was the teacher mode. And then it would, for example, show you answers. You could say, what's the answer to this question. It would say the answers, selection B. And then it would say. Do you want to brainstorm ways of getting your students excited about this topic or of discussing this question with your students? Things like that. So had some pretty cool uses for trying to help teachers. So I am back. This is just a very quick, fast episode because I wanted to just put something out there and say, hi, I am still around. And look forward to, I've got interviews coming up with like I said, vital, which is the doctor. Jargon to patient translator. Plus they do a bunch of other stuff around. Machine learning around medical records and things like that. Also have a discussion with me and a bunch of other podcasters at AI4 that we'll be publishing sometime soon. And then also going to be talking to a bunch of other people that I met at AIA four and other places. One last thing. If you are familiar with tree of thought. I would love to chat with you. There's a contact form on the website, PromptEngineeringPodcast.com or email me, reply in the YouTube comments, I would love to bring someone on to discuss those. Oh, and also if anyone is working on constitutional AI or is familiar with it, I would love to bring someone on to talk about that because I am very curious about it. Thanks. Hope you're having a great week and I will talk to you soon. Bye. Thanks for coming to the prompt engineering podcasts podcast dedicated helping you be a better prompt engineer Episodes are released every Wednesday I also host masterminds where you can collaborate with me and 50 other people live on zoom to improve your prompts Join us at promptengineeringmastermind. com for the schedule of the upcoming masterminds. Finally, please remember to like and subscribe. See you next week.