Superdatascience

In this episode, the Director of Architecture at NVIDIA, Dr. Magnus Ekman, joins Jon Krohn to discuss how machine learning, including deep learning, can optimize computer hardware design. The pair also review his exceptional book 'Learning Deep Learning.' In this episode you will learn: What hardware architects do [10:15] How ML can optimize hardware speed [ 13:19] Magnus’s Deep Learning Book [21:14] Is understanding how ML models work important? [36:16] Algorithms inspired by biological evolution [41:25] How artificial general intelligence won’t be obtained by increasing model parameters alone [51:24] Why there will always be a place for CNNs and RNNs [54:51] How people can "transition" realistically into ML [1:09:15] Additional materials: www.superdatascience.com/575

In this episode, Jon shares how the right music can power your productivity. It's no secret that he's a big fan of 'deep work,' but this week, he opens up about the artists, sites, and playlists that propel his productivity to new levels. Additional materials: www.superdatascience.com/574

In this episode, co-founder and CEO of Linea, Dr. Doris Xin, joins Jon Krohn to discuss how automating ML model deployment delivers groundbreaking change to data science productivity, and shares what it's like being the CEO of an exciting, early-stage tech start-up. In this episode you will learn: How Linea reduces ML model deployment down to a couple of lines of Python code [5:14] Linea use cases [11:30] How DAGs can 10x production workflow efficiency [22:12] ML model graphlets and reducing wasted computation [24:14] What future Doris envisions for autoML [35:23] Doris’s day-to-day life as a CEO of an early-stage start-up [42:43] What Doris looks for in the engineers and data scientists that she hires [52:21] The future of Data Science and how to prepare best for it [53:58] Additional materials: www.superdatascience.com/573

In this episode, Jon shares his habit of blocking out two hours in his mornings that are free from email and social media distractions. Tune in to learn how this habit helps him deeply focus on his most delightful tasks of the day. Additional materials: www.superdatascience.com/572

Einblick co-founder and associate professor at MIT, Tim Kraska, joins Jon Krohn to discuss no-code collaboration tools for data science and uncovers the clever database and machine learning tricks under the hood of the visual data computing platform. In this episode you will learn: The inspiration behind Einblick [2:45] Einblick's progressive approximation engine [6:43] How no-code tools impact productivity [17:18] The critical steps to become more data-driven as an organization [24:30] How research universities like MIT support high-risk, long-term research [38:37] How ML applied to databases enables them to be faster and more efficient [42:03] How real-time collaboration environments like Google Docs are likely to become more widespread for data science tasks [ 49:24] Additional materials: www.superdatascience.com/571

In this episode, Jon is back with another A.I. model breakthrough! He updates listeners on OpenAI's outstanding DALL-E 2 model. The new natural language processing model churns out staggering visual examples of whatever text your mind can dream up. Additional materials: www.superdatascience.com/570

Research Scientist at Meta AI, Dr. Noam Brown, joins Jon Krohn to discuss his award-winning no-limit poker-playing algorithms and the real-world implications of his game-playing A.I. breakthroughs. In this episode you will learn: What Meta A.I. is and how it fits into Meta, the company [3:01] Noam's award-winning no-limit poker-playing algorithms, Libratus and Pluribus algorithms. [4:33] What game theory is and how does Noam integrate it into his models? [8:45] The real-world implications of Noam’s game-playing A.I. breakthroughs [25:24] Why Noam elected to become a researcher at a big tech firm instead of in academia [27:06] The main barriers to getting AI game theory techniques beyond games to self-driving cars [30:16] Recommendations for people who want to break into poker AI [37:45] Additional materials: www.superdatascience.com/569

In this episode, Jon updates listeners on one of the industry's biggest breakthroughs to date –Google's new natural language processing model, PaLM. The key innovation with PaLM is scaling up Google's Pathways modeling approach to half a trillion parameters — many-fold more parameters than had previously been trained using this approach. Additional materials: www.superdatascience.com/568

In this episode, the MIT Press Director and Publisher, Dr. Amy Brand, joins Jon Krohn to discuss open-access publishing in data science and how to address the inequalities that exist for women and minorities in STEM. In this episode you will learn: What it’s like to run the prestigious MIT Press [4:34] How open access makes scholarly work more impactful [6:34] How publishing outstanding STEM books for broader audiences, including for children, can help address STEM biases [19:28] Amy's award-winning documentary Picture A Scientist [25:28] What it's like to executive produce a documentary [37:24] What can be done to change STEM to make it more welcoming to minorities [48:44] The best open-source model going forward [58:26] What fascinates Amy about natural language processing [1:01:30] How author metadata in standardized taxonomies can help authors receive the credit they deserve [1:04:50] Additional materials: www.superdatascience.com/567

In this episode, Jon reflects on the Chinese proverb: "The best time to plant a tree was 20 years ago. The second best time is now." He also challenges listeners to reflect on their long-term goals that have gone unfulfilled. Additional materials: www.superdatascience.com/566

In this episode, Jeremie Harris dives into the stirring topic of AI Safety and the existential risks that Artificial General Intelligence poses to humankind. In this episode you will learn: Why mentorship is crucial in a data science career development [15:45] Canadian vs American start-up ecosystems [24:18] What is Artificial General Intelligence (AGI)? [38:50] How Artificial Superintelligence could destroy the world [1:04:00] How AGI could prove to be a panacea for humankind and life on the planet. [1:27:31] How to become an AI safety expert [1:30:07] Jeremie's day-to-day work life at Mercurius [1:35:39] Additional materials: www.superdatascience.com/565

In this episode, Jon speaks with the CEO of Hugging Face, Clem Delangue, about open-source machine learning and transformer architectures, while attending the ScaleUp:AI Conference in New York. Additional materials: www.superdatascience.com/564

In this episode, superstar data science YouTuber Tina Huang joins us to discuss what it's like to work at one of the world's largest tech companies, her strategies for efficient learning, and how best to prepare for a career in data science from scratch. In this episode you will learn: The key areas to focus on when getting started in data science [6:01] Tina’s five steps to consistently doing anything [11:55] Tina's day-to-day life as a data scientist at one of the world’s largest tech companies [20:02] How Tina's computer science background helps her work [26:20] Traditional banking culture vs big tech [32:12] How Tina's background in pharmacology impacts her work in data science [36:15] The software languages that Tina uses daily in her work [45:30] How Tina’s SQL course practically prepares you for data science interviews [47:24] Additional materials: www.superdatascience.com/563

In this episode, Jon shares his daily technical exercise, which is part of an extensive habit tracking system that allows him to achieve more, create more structure within his day, and cut out bad habits. By completing mathematics, computer science, or programming exercise daily, Jon is able to hone his technical skills in a limitlessly broad field and open new professional opportunities in the long run. Additional materials: www.superdatascience.com/562

In this episode, Ribbon Health CTO Nate Fox joins us to discuss the ins and outs of APIs. Tune in to hear him share how he and his team build out APIs from scratch; how they ensure the uptime and reliability of APIs and how they leverage machine learning to improve the quality of healthcare delivery and maximize their social impact. In this episode you will learn: What are APIs? [13:20] How Ribbon Health’s data API leverages ML models to improve the quality of healthcare delivery [16:08] How to design a data API from scratch [20:00] How to ensure the uptime and reliability of APIs [25:28] How Ribbon uses knowledge graphs, manually labeled data samples, and an XGBoost model with hundreds of inputs to assign a confidence score [27:14] Nate’s favorite tool for easily scaling up the impact of data science [37:40] What is Nate’s day-to-day like? [34:34] The qualities Nate looks for when hiring data scientists [39:50] How scientists and engineers can make a big social impact in health technology [42:50]

In this episode, Jon shares his daily habit of reading two pages and explains how it has transformed his productivity. Additional materials: www.superdatascience.com/560

Natural language processing expert and PhD student Melanie Subbiah sits down with Jon Krohn to discuss GPT-3, its strengths and weaknesses, and the future of NLP. In this episode you will learn: What is GPT-3? [6:24] The strengths and weaknesses of GPT-3 [14:38] What is autoregression? [18:03] GPT-3's new fine-tuning abilities [20:02] Bias issues with GPT-3 [22:47] The future of natural language processing models [27:54] How Melanie ended up working at OpenAI [38:13] Melanie’s self-study process [42:19] Melanie's work on OpenAI API [45:45] How to address the climate change and bias issues that cloud discussions of large natural language models [49:40] Why Melanie chose to do a PhD at Columbia University [1:01:17] The machine learning tools Melanie’s most excited about [1:08:09] Additional materials: www.superdatascience.com/559

In this episode, Jon shares the key topics he recently discussed with the Open Data Science Conference. From the approach behind his extensive machine learning and deep learning content library to revealing the key tools and software he uses daily, get to know Jon and his process a little better. Additional materials: www.superdatascience.com/558

Pandas expert Matt Harrison sits down with Jon Krohn to discuss tips, tricks and best practices for Pandas learning and mastery. In this episode you will learn: Pros and cons of self-publishing and working with a publisher [5:05] Matt's six tips for using Pandas [17:13] The best way for corporate teams to level up their skills [40:04] How to learn anything effectively [47:14] Matt’s tricks for staying motivated [50:00] Matt’s recommendations for using Git and the Unix command line [1:00:14] Matt’s recommended software libraries for working with tabular data [1:19:45] Additional materials: www.superdatascience.com/557

Discover Jon’s extensive library of machine learning content and learn why Jon's Machine Learning House forms the knowledge structure of an outstanding data scientist or ML engineer. Additional materials: www.superdatascience.com/556