127: AI for Code with Eran Yahav
Programming Throwdown
Are You Trying to Implement a Hybrid System?
In the background, half a million lines of rustcod that are running under the hood and doing very efficient neuro network inference. Two to four different models that are being kind of combined together, and have different trade offs in terms of response time and accuracy. If your type bing slightly slower engineer may catch up and you may get slightly better predictions because the stronger model kind of made it in time. And if not, you may be getting results from a slightly inferior model. How do you make predictions at the right places that do not interrupt the human byt a kind of finding the balance?
00:00
Transcript
Play full episode
Remember Everything You Learn from Podcasts
Save insights instantly, chat with episodes, and build lasting knowledge - all powered by AI.
