Super Data Science: ML & AI Podcast with Jon Krohn

AI certification and EU regulation: Jan Zawadzki, CTO and CO Managing Director of Certif.ai, talks to Jon Krohn about the future of certification for AI startups and keeping within rigorous international regulations. Additional materials: www.superdatascience.com/736 Interested in sponsoring a SuperDataScience Podcast episode? Visit JonKrohn.com/podcast for sponsorship information.

Artificial General Intelligence, AlphaGo, and Google DeepMind: Jon Krohn speaks to Mehdi Ghissassi, Director of Product Management at Google DeepMind, about the ethics and social impact of AI, keeping up with AI releases with safety in mind, and other pressing AI problems that keep him awake at night. In this episode, Mehdi and Jon also take a broader look at the current AI landscape, the opportunities for AI investors and startups, and what AI product managers need to get ahead. This episode is brought to you by Gurobi (https://gurobi.com/sds), the Decision Intelligence Leader, and by CloudWolf (https://www.cloudwolf.com/sds), the Cloud Skills platform. Interested in sponsoring a SuperDataScience Podcast episode? Visit JonKrohn.com/podcast for sponsorship information. In this episode you will learn: • How DeepMind seeks to ‘solve intelligence’ [05:14] • The impact of AGI’s capabilities on medicine [16:37] • How the general public might come to apply future AI systems [28:09] • How working on product development for Africa has shaped Mehdi’s perspective on AI’s potential and challenges [37:17] • How to stay on top of rapid changes in AI [39:17] • What investors look for in AI startups [59:16] • Tips for product managers [1:03:34] Additional materials: www.superdatascience.com/735

Robot Soccer takes center stage as Jon Krohn and Dário Catarrinho, Secretary of the Dutch Nao Team and an AI student at the University of Amsterdam, discuss the intricate machine learning that enables robots to navigate the field, make decisions in real-time, respond to sound, and compete against each other in a gripping display of skill and strategy. Additional materials: www.superdatascience.com/734 Interested in sponsoring a SuperDataScience Podcast episode? Visit JonKrohn.com/podcast for sponsorship information.

Yannic Kilcher, a leading ML YouTuber and DeepJudge CTO, teams up with Jon Krohn this week to delve into the open-source ML community, the technology powering Yannic’s Swiss-based startup, and the significant implications of adversarial examples in ML. Tune in as they also unpack Yannic's approach to tracking ML research, future AI prospects and his startup challenges. This episode is brought to you by Gurobi (https://gurobi.com/sds), the Decision Intelligence Leader, and by CloudWolf (https://www.cloudwolf.com/sds), the Cloud Skills platform. Interested in sponsoring a SuperDataScience Podcast episode? Visit JonKrohn.com/podcast for sponsorship information. In this episode you will learn: • About OpenAssistant project [03:39] • Alignment issues in open-source vs closed-source [08:36] • Alternative formulas vital for crafting superior LLMs [20:29] • Strategies to foster open-source LLM ecosystems [27:07] • Yannic's pioneering work in legal document processing at DeepJudge [31:31] • Comprehensive overview of adversarial examples [1:04:02] • The future AI's landscape [1:18:08] • Startup challenges [1:25:35] Additional materials: www.superdatascience.com/733

Exploring our vast universe, in this episode Jon Krohn meets with Daniela Huppenkothen at the University of Amsterdam's astronomy department for a wide-ranging discussion about building instrumentation for telescopes, collecting data from outer space and how to sort astronomy’s problem of enormous amounts of data. Additional materials: www.superdatascience.com/732 Interested in sponsoring a SuperDataScience Podcast episode? Visit JonKrohn.com/podcast for sponsorship information.

Ethics and machine intelligence pioneer Nell Watson speaks to host Jon Krohn about the differences between AI ethics and AI safety, how crying wolf may result in future complications for AI development and the importance of ensuring IEEE standards to mitigate and regulate AI risks. She also touches on what she considers a “second Enlightenment”, in which we may start to form intimate relationships with AI—to both parties’ benefit. This episode is brought to you by Gurobi (https://gurobi.com/sds), the Decision Intelligence Leader, and by CloudWolf (https://www.cloudwolf.com/sds), the Cloud Skills platform. Interested in sponsoring a SuperDataScience Podcast episode? Visit JonKrohn.com/podcast for sponsorship information. In this episode you will learn: • AI ethics and AI safety [05:30] • How "moving fast" could break the world [18:07] • The shifting relationship between humans and machines [29:54] • International ethics standards, and their review process [52:10] • Current and future ethical standards [1:05:31] • Building a universal basic income with AI [1:19:23] Additional materials: www.superdatascience.com/731

In this episode, Kyle Daigle, COO of GitHub, joins Jon Krohn to discuss the transformative impact of generative AI tools like GitHub Copilot. Learn how these tools streamline software development, enhance collaboration, and accelerate code reviews. Discover innovative approaches to collaboration and innersourcing, reshaping the future of teamwork in the digital age. Additional materials: www.superdatascience.com/730 Interested in sponsoring a SuperDataScience Podcast episode? Visit JonKrohn.com/podcast for sponsorship information.

Dr. Blake Richards discusses the world of AI and human cognition this week. Learn about the essence of intelligence, the ways AI research informs our understanding of the human brain, and discover the potential future scenarios where AI and humanity might intersect. This episode is brought to you by Gurobi (https://gurobi.com/sds), the Decision Intelligence Leader, and by CloudWolf (https://www.cloudwolf.com/sds), the Cloud Skills platform. Interested in sponsoring a SuperDataScience Podcast episode? Visit JonKrohn.com/podcast for sponsorship information. In this episode you will learn: • Blake's research and his take on intelligence [09:56] • How we can evaluate progress in artificial general intelligence [15:54] • Blake's thoughts on biomimicry [20:57] • Why Blake thinks the fears regarding AI are overdone [25:38] • The most effective strategies to mitigate AI fears without hindering innovation [35:31] • What steps can we take to ensure that AI supports human flourishing [45:23] • The importance of interpreting neuroscience data through the lens of ML [55:08] • Backpropagation, gradient descent and the brain [1:17:32] Additional materials: www.superdatascience.com/729

Learn how to achieve human-like outputs from LLMs in this week’s Five-Minute Friday with Jon Krohn. Understand the various current methods available to decode and generate text, as well as the differences between them. Find out about greedy search, beam search, sampling, and contrastive search, and how you can use them to create incredibly useful LLMs. Additional materials: www.superdatascience.com/728 Interested in sponsoring a SuperDataScience Podcast episode? Visit JonKrohn.com/podcast for sponsorship information.

Coded bias, intersectionality in AI, and computer vision: Founder of the Algorithmic Justice League Joy Buolamwini talks to host Jon Krohn about the impact of exclusion and inclusion in datasets, the need to address intersectionality when identifying racial, age, or gender-based prejudice in machine learning tools, protections for artists and creative practitioners against AI, and the role that AI may have in combating systemic racism. This episode is brought to you by Gurobi (https://gurobi.com/sds), the Decision Intelligence Leader, and by CloudWolf (https://www.cloudwolf.com/sds), the Cloud Skills platform. Interested in sponsoring a SuperDataScience Podcast episode? Visit JonKrohn.com/podcast for sponsorship information. In this episode you will learn: • What coded bias is [06:49] • The problem with bias in machine learning datasets [18:41] • The Incoding Movement [42:08] • About the Pilot Parliaments Benchmark [52:07] • Ethics and the future of AI [1:20:10] • The potential for AI to end systemic racism [1:32:59] Additional materials: www.superdatascience.com/727

Ben Jones, CEO of Data Literacy, discusses the seven crucial components of effective data leadership. From ethics to technology and fostering a data-centric culture, Jones provides actionable insights and practical examples. Tune in to empower your organization with purposeful and ethical data strategies from day one. Additional materials: www.superdatascience.com/726 Interested in sponsoring a SuperDataScience Podcast episode? Visit JonKrohn.com/podcast for sponsorship information.

Dr. Kim Stachenfeld, Research Scientist at Google DeepMind and Affiliate Professor at Columbia University, delves into the realms of AI and neuroscience as she discusses computer-based simulations of the human brain, the efficiency of language in compression, and the neuroscience theories shaping the future of artificial intelligence. Discover the secrets behind memory formation, cognitive enhancement, and the potential of Artificial General Intelligence (AGI) in this thought-provoking episode. This episode is brought to you by Gurobi (https://gurobi.com/sds), the Decision Intelligence Leader, and by ODSC (https://odsc.com), the Open Data Science Conference. Interested in sponsoring a SuperDataScience Podcast episode? Visit JonKrohn.com/podcast for sponsorship information. In this episode you will learn: • The importance of simulations in the context of human intelligence [05:44] • The basic approach to simulating human intelligence or physical systems [09:30] • Will simulations help us realize AGI? [37:21] • The cross-disciplinary potential of LLMs [40:20] • The special role of our brain’s hippocampus in memory formation [1:05:15] • Kim's research on reinforcement learning and neural representation [1:15:02] • Compression in representation learning [1:38:51] • What skills should an aspiring computational neuroscientist hone [1:50:30] Additional materials: www.superdatascience.com/725

In this Friday episode, host Jon Krohn talks to UCSF’s David Moses about BRAVO (Brain-Computer Interface Restoration of Arm and Voice), a study led by Edward Chang and Karunesh Ganguly that helps patients who have lost the ability to speak to communicate once again via a speech neuroprosthesis. Postdoctoral engineer David Moses, who is a part of BRAVO, reveals the data and machine learning models that help BRAVO predict the words and facial expressions that a paralyzed patient is trying to form via their brain activity, crucially helping patients to communicate with medical practitioners and loved ones. Additional materials: www.superdatascience.com/724 Interested in sponsoring a SuperDataScience Podcast episode? Visit JonKrohn.com/podcast for sponsorship information.

Mathematical optimization should be known to every data scientist: Jon Krohn speaks to Jerry Yurchisin, Data Science Strategist at Gurobi, the decision-making technology and best-kept secret of 80% of America’s leading enterprises. This episode is brought to you by the Zerve data science dev environment (https://zerve.ai), the Decision Intelligence Leader, by ODSC (https://odsc.com), the Open Data Science Conference, and by CloudWolf (https://www.cloudwolf.com/sds), the Cloud Skills platform. Interested in sponsoring a SuperDataScience Podcast episode? Visit JonKrohn.com/podcast for sponsorship information. In this episode you will learn: • What mathematical optimization is [04:27] • How Gurobi solver works [29:01] • How to use Gurobi with Python [36:08] • Coding and algebra resources [41:14] • When to use mathematical optimization and machine learning together [54:23] • Using mathematical optimization in natural language processing [1:01:00] Additional materials: www.superdatascience.com/723

This episode delves into an intriguing research paper from top institutions like UC Irvine and MIT, analyzing the carbon emissions of AI-driven writing and illustrating versus traditional human methods. The findings might surprise you. Is AI the more eco-friendly option? Listen now to explore this compelling intersection of technology and sustainability. Additional materials: www.superdatascience.com/722 Interested in sponsoring a SuperDataScience Podcast episode? Visit JonKrohn.com/podcast for sponsorship information.

Dr. Amira Abbas, Quantum Computing Researcher at the University of Amsterdam, explores the captivating world of Quantum Machine Learning. Learn about the distinct characteristics of qubits and the vital processes of Quantum ML. For those keen on exploring further, Amira offers noteworthy ML tools suggestions to kickstart your journey in Quantum Computing. This episode is brought to you by Gurobi (https://gurobi.com/sds), the Decision Intelligence Leader, by ODSC (https://odsc.com), the Open Data Science Conference, and by CloudWolf (https://www.cloudwolf.com/sds), the Cloud Skills platform. Interested in sponsoring a SuperDataScience Podcast episode? Visit JonKrohn.com/podcast for sponsorship information. In this episode you will learn: • Quantum computing vs classical computing [03:42] • What is quantum entanglement [11:45] • What is a qubit [15:07] • The best problems for quantum ML [30:08] • Three distinct steps in quantum ML and its potential [39:06] • Quantum neural networks [49:03] • What Amira's working on at the moment [1:10:20] • How to get started in quantum ML [1:21:06] • Amira's recommended ML tools for quantum computing [1:30:39] Additional materials: www.superdatascience.com/721

DALL-E may be playing second fiddle to Midjourney no longer with OpenAI’s latest model for generative AI art, DALL-E 3. Host Jon Krohn breaks down the newest model’s capabilities to go beyond producing incredible artistic images, and that follows your written brief to the letter. Additional materials: www.superdatascience.com/720 Interested in sponsoring a SuperDataScience Podcast episode? Visit JonKrohn.com/podcast for sponsorship information.

In this episode, Margot Gerritsen and Jon Krohn discuss the fundamentals of computational mathematics and its application in studying fluid dynamics. Margot also talks about how her synesthesia led to a lifelong interest in math, using computational mathematics to predict airflow, and why it is so important that underrepresented groups in data science become more visible through organizations like Women in Data Science. This episode is brought to you by the Zerve data science dev environment (https://zerve.ai), by Gurobi (https://gurobi.com/sds), the Decision Intelligence Leader, and by ODSC (https://odsc.com), the Open Data Science Conference. Interested in sponsoring a SuperDataScience Podcast episode? Visit JonKrohn.com/podcast for sponsorship information. In this episode you will learn: • About computational mathematics and its relation to data science [03:19] • Margot’s current research into emissions simulation [15:05] • Computational Mathematics: Real-World Applications [33:18] • The importance of wind tunnels in testing designs [47:54] • The beauty of linear algebra [1:05:59] • Synesthesia: Seeing Numbers as Colors [1:16:33] • About Women in Data Science [1:24:59] Additional materials: www.superdatascience.com/719

Elevate your ChatGPT game with a useful custom instruction. Tune in to hear Jon’s trick for maximizing ChatGPT’s potential. Additional materials: www.superdatascience.com/718 Interested in sponsoring a SuperDataScience Podcast episode? Visit JonKrohn.com/podcast for sponsorship information.

Dr. Dan Shiebler, Head of ML at Abnormal Security, joins Jon Krohn this week and unveils the intricacies of cybercrime detection and email protection, and the role of AI in future challenges. This episode is brought to you by Grafbase (https://grafbase.com), the unified data layer, by ODSC (https://odsc.com/), the Open Data Science Conference, and by Modelbit (https://modelbit.com), for deploying models in seconds. Interested in sponsoring a SuperDataScience Podcast episode? Visit JonKrohn.com/podcast for sponsorship information. In this episode you will learn: • The heuristic and “intermediate” ML models that they develop at Abnormal Security [07:08] • How Dan uses LLMs at Abnormal Security [15:46] • How false negatives are individually the biggest classification error to avoid in cybersecurity [20:49] • How head-to-head competitor analysis helps refine models [34:34] • Resilient ML in cybersecurity [38:36] • Abnormal Security’s routine for updating their models [52:37] • AI's impact on the urban world [1:09:57] • How to stay updated in data science and AI [1:13:46] Additional materials: www.superdatascience.com/717