Vanishing Gradients

Hugo and Johno discuss the evolution of tooling and accessibility in AI over the past decade, highlighting advancements in using big models from Hugging Face and hi-res satellite data. They delve into the Generative AI mindset, democratizing deep learning with fast.ai, and the importance of UX in generative AI applications. The discussion also covers the skill set needed to be an LLM and AI guru, as well as efforts at answer.ai to democratize LLMs and foundation models.

Allen Downey discusses statistical paradoxes and fallacies in using data, including the base rate fallacy and algorithmic fairness. They dive into examples like COVID vaccination data and explore the challenges of interpreting statistical information correctly. The conversation also covers topics such as epidemiological paradoxes, Gaussian distributions, and the importance of understanding biases in data interpretation for media consumption.

Jeremy Howard (Fast.ai), Shreya Shankar (UC Berkeley), and Hamel Husain (Parlance Labs) join Hugo Bowne-Anderson to talk about how LLMs and OpenAI are changing the worlds of data science, machine learning, and machine learning engineering.Jeremy Howard (https://twitter.com/jeremyphoward) is co-founder of fast.ai, an ex-Chief Scientist at Kaggle, and creator of the ULMFiT approach on which all modern language models are based. Shreya Shankar (https://twitter.com/sh_reya) is at UC Berkeley, ex Google brain, Facebook, and Viaduct. Hamel Husain (https://twitter.com/HamelHusain) has his own generative AI and LLM consultancy Parlance Labs (https://parlance-labs.com/) and was previously at Outerbounds, Github, and Airbnb.They talk aboutHow LLMs shift the nature of the work we do in DS and ML,How they change the tools we use,The ways in which they could displace the role of traditional ML (e.g. will we stop using xgboost any time soon?),How to navigate all the new tools and techniques,The trade-offs between open and closed models,Reactions to the recent Open Developer Day and the increasing existential crisis for ML.LINKSThe panel on YouTube (https://youtube.com/live/MTJHvgJtynU?feature=share)Hugo and Jeremy's upcoming livestream on what the hell happened recently at OpenAI, among many other things (https://lu.ma/byxyzfrr?utm_source=vg)Vanishing Gradients on YouTube (https://www.youtube.com/channel/UC_NafIo-Ku2loOLrzm45ABA)Vanishing Gradients on twitter (https://twitter.com/VanishingData) This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit hugobowne.substack.com

Guest Hamel Husain, a machine learning engineer, discusses the business value of large language models (LLMs) and generative AI. They cover common misconceptions, necessary skills, and techniques for working with LLMs. The podcast explores the challenges of working with ML software and chat GPT, the importance of data cleaning and analysis, and deploying LLMs in production with guardrails. They also discuss an AI-powered real estate CRM and optimizing marketing strategies through data analysis.

Chris Wiggins, Chief data scientist for the New York Times, and Matthew Jones, professor of history at Princeton University, discuss their book on the history of data and its impact on society. They explore topics such as the use of data for decision making, the development of statistical techniques, the influence of Francis Galton on eugenics, and the rise of data, compute, and algorithms in various fields.

Hugo chats with Katharine Jarmul, a Principal Data Scientist at Thoughtworks Germany, specializing in privacy and ethics in data workflows. They dive into the vital distinctions between data privacy and security, demystifying common misconceptions. Katharine highlights the impact of GDPR and CCPA, and explores advanced concepts like federated learning and differential privacy. They also tackle real-world issues like privacy attacks and the ethical responsibilities of data scientists, making a compelling case for prioritizing privacy in data practices.

Eric Ma, a leader in the research team at Moderna Therapeutics, discusses the tools and techniques used for drug discovery, the importance of machine learning and Bayesian inference, and the cultural questions surrounding hiring and management in research data science in biotech. They also explore the tech stack used in their work, the skills and hiring considerations in biotech, the importance of data testing and standardizing Excel spreadsheets, and the current state and challenges of Bayesian inference.

Hugo speaks with Tanya Cashorali, a data scientist and consultant that helps businesses get the most out of data, about what end-to-end data science looks like across many industries, such as retail, defense, biotech, and sports, includingscoping out projects,figuring out the correct questions to ask,how projects can change,delivering on the promise,the importance of rapid prototyping,what it means to put models in production, andhow to measure success.And much more, all the while grounding their conversation in real-world examples from data science, business, and life.In a world where most organizations think they need AI and yet 10-15% of data science actually involves model building, it’s time to get real about how data science and machine learning actually deliver value!LINKSTanya on Twitter (https://twitter.com/tanyacash21)Vanishing Gradients on YouTube (https://www.youtube.com/channel/UC_NafIo-Ku2loOLrzm45ABA)Saving millions with a Shiny app | Data Science Hangout with Tanya Cashorali (https://youtu.be/qdAroyFRFCg)Our next livestream: Research Data Science in Biotech with Eric Ma (https://www.eventbrite.com/e/research-data-science-in-biotech-tickets-550400882857?aff=fs) This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit hugobowne.substack.com

JD Long, agricultural economist and quant, discusses decision making under uncertainty in data science, common mistakes, heuristics for decision-making, and the impact of cognitive biases. Topics include coupling data science with decision-making, model building, storytelling, and the intersection of cognitive biases.

Hugo speaks with JD Long, agricultural economist, quant, and stochastic modeler, about decision making under uncertainty and how we can use our knowledge of risk, uncertainty, probabilistic thinking, causal inference, and more to help us use data science and machine learning to make better decisions in an uncertain world. This is part 1 of a two part conversation. In this, part 1, we discuss risk, uncertainty, probabilistic thinking, and simulation, all with a view towards improving decision making and we draw on examples from our personal lives, the pandemic, our jobs, the reinsurance space, and the corporate world. In part 2, we’ll get into the nitty gritty of decision making under uncertainty.As JD says, and I paraphrase, “You may think you train your models, but your models are really training you.”LinksVanishing Gradients' new YouTube channel! (https://www.youtube.com/channel/UC_NafIo-Ku2loOLrzm45ABA)JD on twitter (https://twitter.com/CMastication)Executive Data Science, episode 5 of Vanishing Gradients, in which Jim Savage and Hugo talk through decision making and why you should always be integrating your loss function over your posterior (https://vanishinggradients.fireside.fm/5)Fooled by Randomness by Nassim Taleb (https://en.wikipedia.org/wiki/Fooled_by_Randomness)Superforecasting: The Art and Science of Prediction Philip E. Tetlock and Dan Gardner (https://en.wikipedia.org/wiki/Superforecasting:_The_Art_and_Science_of_Prediction)Thinking in Bets by Annie Duke (https://www.penguin.com.au/books/thinking-in-bets-9780735216372)The Signal and the Noise: Why So Many Predictions Fail by Nate Silver (https://en.wikipedia.org/wiki/The_Signal_and_the_Noise)Thinking, Fast and Slow by Daniel Kahneman (https://en.wikipedia.org/wiki/Thinking,_Fast_and_Slow) This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit hugobowne.substack.com