What's AI Podcast by Louis-François Bouchard

Here's an interview with the Director of Perception at Zoox, Ruijie (RJ) He, with the goal of demystifying what is a good profile to get an ML engineer job and perform at the interviews. We go over many points from the current job market, the screening process, how to prepare for an interview, the interview process and the role of an ML engineer at Zoox. For a quick introduction to this great company: Zoox is an artificial intelligence company focused on a single product: autonomous vehicles. They are building intelligent taxis for the future hiring a lot of people in the whole machine learning / data science space. ►Learn more about Zoox: https://zoox.com/ ►Support me on Patreon: https://www.patreon.com/whatsai Chapters: 0:34 Who is RJ He and what is Zoox 1:14 What role is Zoox looking for? 2:10 What is an ML Engineer? 4:48 From research to a product, what is the process at Zoox? 7:01 What are you looking for in ML Engineers? 9:50 The screening process at Zoox 12:52 The interviews at Zoox 14:08 How to get a job without work experience? 16:18 What kind of projects are you looking for? 18:42 Are research publications important/necessary? 19:58 Online certifications vs. university 21:32 The interview process (then vs. now) 22:55 The AI field and the job market (then vs. now) 24:42 Hiring at Zoox, future interviews and conclusion

Here's an interview with Adam Grzywaczewski, the senior deep learning architect at NVIDIA. In this interview, we talk about his role, and the interview process to get into such a role, and we learn more about NVIDIA and more interesting insights from Adam. [FOR THE 4080 GIVEAWAY] Comment under this video and send me a screenshot DURING GTC to enter the RTX 4080 (and 10 DLI credits) giveaway! Adam's GTC events: The Possibilities for Natural Language Processing and Large Language Models in Finance: Insights from Deutsche Bank [S51160]: https://www.nvidia.com/gtc/session-catalog/?ncid=ref-inpa-477072&?tab.catalogallsessionstab=16566177511100015Kus&search=#/session/1666078863989001bkfy Connect with the Experts: Deep Learning, Machine Learning, and Data Science [CWES52118]: https://www.nvidia.com/gtc/session-catalog/?ncid=ref-inpa-477072&?tab.catalogallsessionstab=16566177511100015Kus&search=#/session/1670255843552001iaMr ►My Newsletter (A new AI application explained weekly to your emails!): https://www.louisbouchard.ai/newsletter/ ►Support me on Patreon: https://www.patreon.com/whatsai ►Support me through wearing Merch: https://whatsai.myshopify.com/ ►Join Our AI Discord: https://discord.gg/learnaitogether Chapters: 0:00 Hey! Tap the Thumbs Up button and Subscribe. You'll learn a lot of cool stuff, I promise. 00:58 Academic background 07:55 You were trying to scale models, but the hardware didn’t allow you? 09:25 Did you really want a Ph.D. or was it just to work on the project you had in mind? 10:14 Did you already have a goal in mind? Like getting a good job 12:41 How are you assessing the candidate’s capabilities before and during the interview? 14:43 If you have a lot of resumes, are there any projects or degrees that are more interesting than others? 17:41 What is the shape and format of the interview? 20:08 What is a deep learning architect? 21:55 Other than scaling. In what areas are you working on? 23:36 Are you part of a team that supports companies using Nvidia’s products? 24:40 Could you go over the details of a specific project you’ve had? 26:41 Which complicated challenges require your help? 28:15 How do people that work with you deal with large models or data sets? 30:47 So the current challenge is mainly to find which tool to use and how to do it in a cost-effective way? 32:30 Will you talk in GTC about how to scale and deploy NLP models? 33:30 What is your day-to-day like at Nvidia? 37:20 Would you say that AI technology is now more insane than it was in 2017? 38:10 How do you keep up with this fast rate of progress? 39:08 As the field is maturing, would you say that you have to be more specific on what you’re doing compared to 5-6 years ago? 40:06 Is the need for specific knowledge more challenging than when you had to have broader knowledge? 42:54 What is your favorite tool to use? 43:04 What internal tools are you using? 47:00 Are you surprised by the fact that open-source technologies are so powerful? 50:16 What is the biggest challenge in just deploying models? 53:33 So the main challenges come with the complexity of the solution and the randomness?

If you thought ChatGPT was good, wait before you try this one… GPT-4 may be the most hyped language model we’ve had, with tons of rumors and news even prior to its announcement, and it wasn’t for nothing. Tl;dr: GPT-4 is ChatGPT’s big brother. References:  ►Read the full article: https://www.louisbouchard.ai/gpt-4/  ►Try GPT-4 now: https://chat.openai.com/  ►API waitlist: https://openai.com/waitlist/gpt-4  ►OpenAI blog post: https://openai.com/product/gpt-4  ►GPT-4 research: https://openai.com/research/gpt-4 ►ChatGPT video: https://youtu.be/AsFgn8vU-tQ  ►What is prompting: https://youtu.be/pZsJbYIFCCw  ►Learn Prompting: https://learnprompting.org/  More from me...  ►My Newsletter (A new AI application explained weekly to your emails!): https://www.louisbouchard.ai/newsletter/  Be the first to hear about the coolest models like ChatGPT, and follow my favorite daily AI newsletter: https://www.syntheticmind.io/subscribe?ref=EFowuebnlZ  ►Support me on Patreon: https://www.patreon.com/whatsai ►Support me through wearing Merch: https://whatsai.myshopify.com/  ►Join Our AI Discord: https://discord.gg/learnaitogether  #openai #chatgpt #gpt4

An interview with one of the best Kaggler out there, Chris Deotte. Chris is a Senior Data Scientist at NVIDIA and is getting paid for his Kaggle skills! Kaggle is a platform mainly known for hosting machine learning competitions... Comment under the YT video and send me a screenshot DURING GTC to enter the RTX 4080 giveaway: https://youtu.be/NjGnnG3evmE ►Follow my favorite daily AI newsletter: https://www.syntheticmind.io/subscribe?ref=EFowuebnlZ  ►Support me through wearing Merch: https://whatsai.myshopify.com/   Chris's GTC events:  ►Developing State-of-the-Art Models in a Short Time:  https://www.nvidia.com/gtc/session-catalog/?ncid=ref-inpa-477072&?tab.catalogallsessionstab=16566177511100015Kus&search=#/session/1666650462301001Ltpf ►Learn How to Create Features from Tabular Data and Accelerate your Data Science Pipeline:  https://www.nvidia.com/gtc/session-catalog/?ncid=ref-inpa-477072&?tab.catalogallsessionstab=16566177511100015Kus&search=#/session/1666168670726001zds5 More... ►My Newsletter: https://www.louisbouchard.ai/newsletter/ ►Support me on Patreon: https://www.patreon.com/whatsai ►Join Our AI Discord: https://discord.gg/learnaitogether Chapters: 00:49 What is your academic background?  01:20 How did you get into data science from a mathematics background?  02:04 What is a data scientist for you, and what is your role as one?  02:33 Do you think data science is mainly a role for academia because it’s a lot of statistical and math knowledge? Do you think a PHD or a  masters is necessary to get such a role?  03:47 What is your role as a data scientist at Nvidia?  05:40 What is Kaggle, and what is a grand master at Kaggle?  08:20 Do you think Kaggle competitions are a good way of improving your resume and build experience if you want to work in the industry?  11:54 Is there something specific to Kaggle that doesn't work in the real world?  16:29 Are most competitions similar to one another? Or are there different challenges depending on the competition?  18:34 So Kaggle will allow you to be a generalist?  19:08 What tips would you give to a beginner who wants to participate in the competition and have a chance of winning?  20:43 Do you participate in competitions of every field?  24:17 What is a Kaggle grandmaster and what does it mean to have this four times?  27:52 Was there a category that was harder for you? Or one that you didn't enjoy?  30:38 What was the main factor for Nvidia to find you and hire you?  32:11 How was the interview process if they already knew how you worked and your knowledge?  35:07 How did you prepare for these interviews?  36:28 How can they assess your skills if there are so few people that do what you do?  37:27 Since the technical interviews are in different fields, is it over if you fail one of them?  40:04 Can you describe your day to day at Nvidia?  41:29 So you're being paid to do what you love to do?  43:03 Could you enter into the details of a recent project?  46:10 How do you deal with a very large data set?  48:39 Do you have a machine or are you connected to servers?  49:56 What would you recommend to someone who has a basic laptop and wants to practice DS?  53:37 Do you sometimes need to do particular processes to make it work with multiple GPU's?  56:39 What are the daily tools you use to do data science and Kaggle?  58:00 Is there anything we can learn from Nvidia coming soon?  58:58 Is it accessible for someone just starting at Kaggle?

What is Data Science like at NVIDIA? An interview with Meriem Bendris, Senior Solution Architect at NVIDIA. Sign up to Meriem's free session: https://www.nvidia.com/gtc/session-catalog/?ncid=ref-inpa-477072&tab.catalogallsessionstab=16566177511100015Kus&search=meriem#/session/1670255843552001iaMr The interview answers the questions... 00:00 Hey! Give a Thumbs up to the video If you enjoy it and let me know who or which role you’d like me to interview next!  00:50 How did you get into NVIDIA? What’s your academic background?  04:12 How were the NVIDIA  interviews?  05:54 How did you prepare for the interviews?  09:13 What is a solution architect at Nvidia?  13:47 How are the rôles responsibilities at NVIDIA?  17:15 Do you see any resemblance between your work at NVIDIA and when you were doing your PhD or postgraduate degree?  23:10 When making models more efficients (quantizing), do you reduce performance significantly or do you manage to make them more efficient without sacrificing performance?  25:10 What do you mean by distributing a model and why would you do that?  29:43 Would you say that your PHD was worthwhile?  33:25 How can someone coming from a completely different field make the transition into data science?  40:00 Would you recommend diving into resource usage/management when learning AI?  43:00 What material/hardware do you need when wanting to learn AI?