About Alan

AI Consultant
Technical highlights document • Media pack • Headshots

Short bio

Dr Alan D. Thompson advises world leaders and Fortune 500s on human and artificial intelligence, specializing in large language models and superintelligence. A former chairman for Mensa’s gifted families, Alan authored six books (including one sent to the moon), created the acclaimed 2021 Leta AI project, and published influential papers informing AI strategy at Microsoft, Apple, Bloomberg, the US Government, and the G7. He continues to deliver AI analysis and keynotes worldwide.

Long bio (show footnotes)

Dr Alan D. Thompson brings decades of human intelligence research into the field of artificial intelligence. Alan began exploring the AI field in 1991, with the development of an early AI chatbot using QBasic. He completed his Bachelor of Science (Computer Science, AI, and Psychology) at Edith Cowan University, 2004; studied Gifted Education at Flinders University, 2017; became a Fellow of the Institute of Coaching affiliated with Harvard Medical School, 2017; and received his doctorate from Emerson, 2021. His dissertation was adapted into a book featuring Dr Rupert Sheldrake, Connected: Intuition and Resonance in Smart People.

As national coordinator for Australian Mensa’s gifted families, he designed and developed several high-profile information initiatives,¹ including testing advocacy,² and the 2015 national survey of gifted children.³ In this role and for GE’s award-winning Decoding Genius series,⁴ Alan worked closely with the pioneering human intelligence researcher Prof Miraca Gross. He also appeared for Warner Bros. in Child Genius, for ABC in Making Child Prodigies, for 60 Minutes, and for ABC Catalyst. He has published six books⁵ on intelligence and high performance, and a copy of his bestselling book Bright was sent to the moon aboard Peregrine Mission One.⁶

Across two terms as chairman for Mensa International’s gifted families committee, he facilitated best practice among 54 member countries. After his polarizing article ‘The new irrelevance of intelligence’⁷ was published in the Journal of Australian Mensa in 2020, Alan left the organization and rejoined the AI field. He delivered a live version of the paper to the World Gifted Conference in 2021.

Shortly after the release of OpenAI’s GPT-3 davinci in 2020, Alan developed a manual integration between the large language model and a humanlike avatar. Named after Leta Hollingworth, the founder of gifted intelligence research, the Leta AI series⁸ attracted international attention with over five million views. It is now part of the Internet Archive.⁹ He live-streamed the release of OpenAI’s ChatGPT, and made The ChatGPT prompt book available for free to the public.¹⁰

His bestselling monthly analysis, The Memo,¹¹ is sent to more than 10,000 subscribers in 147 countries, including the majority of Fortune 500 companies. He has consulted to several trillion-dollar organizations, many under ‘no identify’ clauses. His Models Table¹² is the most complete available database of large language model highlights, detailing hundreds of models, beginning with the 2017 Google Transformer.

In 2022, Alan published ‘What’s in my AI?’,¹³ a comprehensive paper on datasets used to train GPT and related models, followed by independent papers on the Google Pathways architecture,¹⁴ Gemini,¹⁵ GPT-5,¹⁶ and Grok¹⁷ models.

Parallel to the 2023 release of OpenAI’s GPT-4 (trained on English and specialized Icelandic cultural data),¹⁸ he was invited by the Icelandic Center for Artificial Intelligence (ICAI) to design the inaugural Executive Master in Artificial Intelligence.¹⁹

In March 2024, Alan released the pioneering documents ‘The Declaration on AI Consciousness’ and the ‘Bill of Rights for AI,’²⁰ well before the concepts were widely accepted by the public.

On The Joe Rogan Experience, he is described as a ‘hyper-optimist’.²¹ As an independent researcher unaffiliated with any institution, his work continues to be widely referenced by Microsoft,²² Apple,²³ BloombergNEF,²⁴ and the US Government.²⁵ In 2024, his GPT-5 paper was included in the G7 (Group of Seven, global economic policy) report in their AI document²⁶ building on The Hiroshima Framework.

He continues to advise government and enterprise on post-2020 AI, delivering keynotes to audiences worldwide, including in Israel, Ukraine, and China.
■