Weights & Biases helps AI developers build better models faster. Quickly track experiments, version and iterate on datasets, evaluate model performance, reproduce models, and manage your ML workflows end-to-end.
Please subscribe to our channel and turn on notifications for future videos!
The models you will build in the future will only be super intelligent but consciousness will never appear there. The biological world in which man lives has a fractal structure, and for humans one of the elements for receiving this information is the thalamus. If, however, technology goes so far that we are able to print similar structures in our brains, it will probably be possible to connect them to the body of a robot, which will acquire consciousness. This is the only way otherwise there is no way to connect to the so-called mother nature using computer code.
Give this man a nobel. Half of the master's students in AI struggle in understanding when seeing too many indices. It's better to demonstrate indices right on the code rather than in notations
It was amazing, I knew that math was involved in ML but I didn't realize how deeply it was connected xD I had a really hard time trying to understand each subject and tbh I didn't fully understand everything way, but I learned a lot about maths and its implication in ML Thanks for the material ♥
Amazing video. I don’t have a strong background in maths, so putting things in terms of programming really helped me start to comprehend this subject. Thank you so much.
Hi Joe and Thomas, I’m an experienced Network Automation Engineer transitioning into AI, focusing on Generative AI, Prompt Engineering, model training, and AI agent automation. I’ve joined several communities and courses, but they often lack actionable guidance, focusing more on keeping learners enrolled. Your expertise in fine-tuning models and overcoming resource constraints really stood out to me. I’m committed to mastering these skills and would greatly appreciate any advice or resources you could share to help me navigate this journey effectively. Thanks for your time!
Essa quadrilha tem um zoológico de animais. Pra soltar nas estradas quando a pessoa perseguida por eles estão passando, pra acidentar as pessoas. Eles usa todos os meios pra tirar a vida da uma pessoa eles eliminam a pessoa jogam na beira da estrada pra fingir de acidente, e usam animais pra colidir com carros e motos quando estão passando e eles eliminam também as pessoas e jogam culpa nos animais.e outros meios de forma que eles nunca apareça na história pra eles continuar praticando outros crimes.
Between these guys and Groq, it's hard to get excited about them when I can't use them in a production environment. Groq's API is useless with their inference use limits. Oh well, I suppose we'll get there eventually.
Cerebras inference is indeed impressive. I was getting 1800 t/s yesterday which is incredible. It is is also incredibly difficult to manage. Utilising all that output is like trying to drink from a fire hose at the moment! Could either of you recommend an agentic set up that I can use in conjunction with Cerebras as a base to build on, for the Metaculus forecasting tournaments?
Why does anyone believe these claims? There is no evidence whatsoever that they are true. He provided no technical rationale as to how he has been able to achieve what the greatest minds and unlimited amounts of capital have been unable to do. Other than "BIG CHIP GOOD". There is no credible third party verification of the performance of Cerebras' chips. And of course the real story is that the company is about to IPO. He is going to raise capital and then try and actually deliver this wonderful chip to the market. The stock will do really well (will look 'so cheap' compared to NVIDIA). He and his fellow employees and existing investors will sell 20-40% of their stock into the market (not all). They will want to have the appearance of being aligned. Years will go on, capital will continue to get invested in this revolutionary chip. It will never be delivered, it will always 'be around the corner'. Then one day we will find they chip cannot achieve what he claims. But the failure will be blamed on NVIDIA because of their market dominance. You did a great job in maintaining your respect during this interview. It must have been frustrating listening to him. This was extremely difficult to watch. But you should have asked some hard questions. You tried at the start, but he got aggressive - so I get it. It would be great to see this CEO interviewed by industry experts.
DEI hire babbling. "We think computers can be a powerful tool. Especially with applications, and fingers. Fingers are very important for to typing be."