I’ve recently been getting into programming for work. Which of course has got me trying to learn more about AI. I’m still a super amateur. I started with bing AI cause that free gpt4 access and bypassing the 4k character limit by telling bing not to reply until I’m done. Now I’m primarily using free plan of phind and its vscode extension. Which I really like. I don’t use the buttons that write code as I always write code myself and rarely copy paste. Helps me learn. I just like that it’s quicker without crazy character limits.
I did recently see that you can pay for bing AI in VS code. Then it got me thinking if I’m gonna pay then I want to make sure it is for something good and prob not Microsoft. So considering paying for phind but now I am trying to look into what others are doing. There’s a world of products I’ve never heard of like hugging face. So I guess it’s time to start my journey on finding which one works best for me. I’ve recently degoogled myself so haven’t really touched bard but if Gemini proves worth it I may.
Just rambling thoughts from a self taught noob programmer.
This is the best summary I could come up with:
Alongside its Gemini generative AI model, Google this morning took the wraps off of AlphaCode 2, an improved version of the code-generating AlphaCode introduced by Google’s DeepMind lab roughly a year ago.
AlphaCode 2 can understand programming challenges involving “complex” math and theoretical computer science.
And, among other reasonably sophisticated techniques, AlphaCode 2 is capable of dynamic programming, explains DeepMind research scientist Rémi Leblond in a prerecorded video.
Dynamic programming entails simplifying a complex problem by breaking it down into easier sub-problems over and over; Leblond says that AlphaCode 2 knows not only when to properly implement this strategy but where to use it.
According to the whitepaper, AlphaCode 2 requires a lot of trial and error, is too costly to operate at scale and relies heavily on being able to filter out obviously bad code samples.
“One of the things that was most exciting to me about the latest results is that when programmers collaborate with [AlphaCode 2 powered by] Gemini, by defining certain properties for the code to follow, the performance [of the model] gets even better,” Collins said.
The original article contains 567 words, the summary contains 181 words. Saved 68%. I’m a bot and I’m open source!
