3

Jeremy Howard

Jeremy Howard is an Australian data scientist, entrepreneur, and educator, known for his pioneering work in deep learning and AI accessibility. He co-founded fast.ai and created the influential ULMFiT algorithm.
Jeremy Howard
Nationality
Australian
Residence
San Francisco, California, United States
Occupation
Data Scientist, Entrepreneur, Educator, Co Founder
Known for
Co-founding fast.ai, ULMFiT algorithm, AI advocacy, founding Enlitic, Masks4All movement, contributions to deep learning accessibility
Accolades
Enlitic ranked in MIT Tech Review’s Top 50 Smartest Companies TED talk with over 2.5 million views Recognition for Masks4All movement
Education
BA in Philosophy (University of Melbourne), Executive Education, Global Leadership and Public Policy ( Harvard Kennedy School)
Social Media
Summary

Jeremy Howard is an Australian data scientist, entrepreneur, and educator, known for his work in deep learning and AI. He is the co-founder of fast.ai, a research institute dedicated to making deep learning more accessible. Alongside Dr Rachel Thomas, Jeremy created the widely-used course, Deep Learning for Coders, and developed the fastai software library, which is one of the most popular deep learning frameworks. His book Deep Learning for Coders with fastai and PyTorch is highly rated, with Google’s Director of Research, Peter Norvig, praising its depth.

Jeremy's notable contributions to AI include the ULMFiT algorithm, a foundational method for training large language models (LLMs), which has been influential in the development of modern models like GPT and ChatGPT. He co-founded Answer.AI with Eric Ries, an AI R&D lab focused on creating practical end-user products based on research breakthroughs.

Jeremy was also the founding CEO of Enlitic, where he applied deep learning to medical diagnostics. Under his leadership, Enlitic was ranked among the world’s smartest companies by MIT Tech Review. He also served as President and Chief Scientist of Kaggle, a platform for data science competitions, where he ranked first globally in 2010 and 2011.

Jeremy has a diverse entrepreneurial background. He co-founded FastMail, an email service, and Optimal Decisions Group, a company focused on insurance pricing optimisation. Both ventures were later acquired by larger companies. He has also been actively involved in public advocacy during the COVID-19 pandemic, co-founding the global Masks4All movement and leading a major evidence review on mask efficacy, published in the Proceedings of the National Academy of Science.

Academically, Jeremy holds roles as an honorary professor at the University of Queensland and a Digital Fellow at Stanford University. He has been involved with institutions like CSIRO’s Data61, where he advises on AI research and development, and Singularity University, where he taught data science and machine learning.

Jeremy's work has been featured in numerous high-impact publications. He has contributed to The Economist, The Guardian, The Washington Post, and many others. His TED talk on the implications of computers that learn has garnered over 2.5 million views. Jeremy continues to mentor startups, contribute to open-source projects, and focus on expanding the reach and impact of AI technologies.

Biography

Jeremy Howard was born on 13 November 1973 in London, United Kingdom. At the age of three, he moved with his family to Melbourne, Australia, where he would later attend Melbourne Grammar School. His academic journey continued at the University of Melbourne, where he pursued a BA in Philosophy. However, Jeremy has openly mentioned that his formal education was less influential than his practical work experiences, as he worked full-time while studying, only attending exams.

Jeremy's career began in management consulting, spending eight years working with prominent firms like McKinsey & Co and AT Kearney. During this time, he focused on business strategy, helping corporations deal with market challenges and regulatory issues. Jeremy was known for his analytical abilities, being one of three global Analytical Specialists at McKinsey & Co, supporting teams with complex data analysis.

In 1999, Jeremy made a significant leap into entrepreneurship by co-founding two companies: FastMail, an early email service offering advanced features like synchronised email, and Optimal Decisions Group, a company that revolutionised insurance pricing by introducing profit-optimised pricing models. Both ventures were successful, and after nearly a decade, FastMail was sold to Opera Software, and Optimal Decisions Group was acquired by ChoicePoint.

In 2010, Jeremy shifted his focus to data science, becoming involved with Kaggle, a platform for predictive modelling competitions. He quickly rose to prominence, ranking as the top participant in global machine learning competitions in 2010 and 2011. His achievements led to his appointment as President and Chief Scientist of Kaggle. During his time there, Jeremy significantly contributed to Kaggle’s growth, developing its competition platform and raising Series A funding. He left Kaggle in 2013.

In 2014, Jeremy founded Enlitic, a pioneering company applying deep learning to medical diagnostics. His work at Enlitic aimed to make medical diagnostics faster, more accurate, and more accessible by using machine learning to analyse medical data. Under his leadership, Enlitic raised $15 million in two years and was ranked by MIT Tech Review as one of the top 50 smartest companies in the world, ahead of major firms like Facebook and SpaceX.

Jeremy’s contribution to deep learning reached new heights with the co-founding of fast.ai in 2016. Together with Dr Rachel Thomas, Jeremy launched the fast.ai initiative to make deep learning more accessible. This led to the development of the fastai library, which has become one of the most popular deep learning frameworks. He also created the Deep Learning for Coders course, the world’s longest-running AI course, and co-authored the book Deep Learning for Coders with fastai and PyTorch, a widely praised resource for deep learning education.

In 2018, Jeremy and his team introduced the ULMFiT algorithm, a groundbreaking advancement in training large language models. This algorithm was foundational in the development of modern AI models, including ChatGPT and similar systems. Jeremy's work on ULMFiT has had a lasting impact on natural language processing.

During the COVID-19 pandemic in 2020, Jeremy played a crucial role in public health advocacy, co-founding the Masks4All movement. His efforts in advocating for widespread mask use, backed by scientific evidence, were published in The Washington Post and Proceedings of the National Academy of Science. His paper on mask efficacy became one of the most viewed papers on preprints.org.

In addition to his entrepreneurial ventures, Jeremy has held academic roles. Since 2022, he has been an honorary professor at the University of Queensland, teaching and researching deep learning. He has also held a Digital Fellow role at Stanford University since 2023 and serves as an Adjunct Science Fellow at CSIRO’s Data61, advising on AI research and development.

Throughout his career, Jeremy has published several research papers and articles. His contributions include influential works on deep learning, such as “Universal Language Model Fine-Tuning for Text Classification” (2018), which has been cited over 5000 times. His TED talk, “The Wonderful and Terrifying Implications of Computers That Learn,” has garnered over 2.5 million views, highlighting his influence in the field of AI.

Currently, in 2024, Jeremy is the co-founder and CEO of Answer.AI, an AI R&D lab focused on developing practical products based on foundational research. He continues his work at fast.ai, where he remains committed to expanding the accessibility of deep learning technologies. Jeremy is also an active mentor for startups, contributing to the development of AI tools and frameworks that have global applications across various industries.

Vision

Jeremy Howard’s vision is to make artificial intelligence, particularly deep learning, accessible to everyone, regardless of their background or technical expertise. He believes that AI has the potential to solve many of the world’s most pressing challenges, especially in fields like medicine, education, and agriculture. By creating easy-to-use tools, practical applications, and educational resources through initiatives like fast.ai, Jeremy aims to empower people from all walks of life to harness the power of AI. His work focuses on bridging the gap between cutting-edge research and real-world applications, making AI technology more inclusive and impactful.

Recognition and Awards

Jeremy Howard has received significant recognition for his contributions to artificial intelligence and deep learning. He co-developed the ULMFiT algorithm, which has had a major impact on modern AI. His work with Enlitic earned the company a spot on MIT Tech Review’s list of the world’s top 50 smartest companies. Jeremy’s advocacy during the COVID-19 pandemic through the Masks4All movement gained global attention, with his paper on mask efficacy becoming the most-read on preprints.org. His TED talk on the implications of AI has been viewed over 2.5 million times, further establishing his influence in the AI community.

References
Jeremy Howard
Nationality
Australian
Residence
San Francisco, California, United States
Occupation
Data Scientist, Entrepreneur, Educator, Co Founder
Known for
Co-founding fast.ai, ULMFiT algorithm, AI advocacy, founding Enlitic, Masks4All movement, contributions to deep learning accessibility
Accolades
Enlitic ranked in MIT Tech Review’s Top 50 Smartest Companies TED talk with over 2.5 million views Recognition for Masks4All movement
Education
BA in Philosophy (University of Melbourne), Executive Education, Global Leadership and Public Policy ( Harvard Kennedy School)
Social Media