As someone with an apparent greater understanding than me of statistics, linear algebra or just math in general judging from what I can see from halfway into the video, what do you think the current Math/CS/Stats undergrads of today should be doing to position themselves to be immensely valuable in the next 2-3 years when we inevitably have better LLMs and an AI dominant landscape ? I'm halfway done with my undergrad in CS and Math, and I feel clueless. It seems like the blow is softened on people with extensive work experience and/or people already way ahead into their Masters/PhD's but what about those of us that are either too early in the pipeline and/or those that are yet to break into the industry? If you had any words on how to adapt, that'd be helpful.
I am fairly certain that whatever you've learned in CS and Maths is a very good foundation & then the best way to start expanding your knowledge would be with some implementations of simpler models, building up the models' complexity (Andrej Karpathy has some nice vids on how to build GPT). I can imagine this field seems very impressive when you look at all the advancements happening & their speed, however when you go into the actual implementations you will see that often it is relatively simple. Feel free to email me if you want to chat more.
Very nice discussion! Thank you 😊
As someone with an apparent greater understanding than me of statistics, linear algebra or just math in general judging from what I can see from halfway into the video, what do you think the current Math/CS/Stats undergrads of today should be doing to position themselves to be immensely valuable in the next 2-3 years when we inevitably have better LLMs and an AI dominant landscape ?
I'm halfway done with my undergrad in CS and Math, and I feel clueless. It seems like the blow is softened on people with extensive work experience and/or people already way ahead into their Masters/PhD's but what about those of us that are either too early in the pipeline and/or those that are yet to break into the industry? If you had any words on how to adapt, that'd be helpful.
I am fairly certain that whatever you've learned in CS and Maths is a very good foundation & then the best way to start expanding your knowledge would be with some implementations of simpler models, building up the models' complexity (Andrej Karpathy has some nice vids on how to build GPT). I can imagine this field seems very impressive when you look at all the advancements happening & their speed, however when you go into the actual implementations you will see that often it is relatively simple. Feel free to email me if you want to chat more.