London Futurists – Details, episodes & analysis

This episode continues our investigation into the potential wide-ranging implications of advanced AI for economics.

Traditionally, value is said to be created by a combination of capital, which covers the cost of materials and equipment, and labour, whereby humans exercise skills, ingenuity, diligence, attention, and more. What has been a constant debate is the appropriate division of rewards between capital and labour. Critics of the operation of capitalism have often predicted that an accumulation of value within small groups of owners of capital will cause economic instabilities and a subsequent collapse. Despite these forecasts, capitalism has, so far, demonstrated great resilience, defying predictions of its collapse. But if human labour is increasingly displaced by advanced automation, the balance of labour and capital will be fundamentally changed, and capitalism will come under unprecedented pressures.

That’s the thesis of our guest today, Ted Shelton. David first met Ted about 25 years ago, when Ted was Chief Strategy Officer of the software development tools company Borland, and David was an executive within the early smartphone industry. Since that time, Ted has worked for a variety of companies in and around Silicon Valley, including PwC, Cognizant Technology Solutions, Catalytic, Bain, and Inflection AI. Recently, he has been giving a great deal of thought to where AI is taking the economy.

Selected follow-ups:

• Ted Shelton's posts on LinkedIn
• "On the Transformation of Capitalism's Fundamental Assumptions Under Conditions of Scaling Machine Intelligence" - working paper by Ted Shelton
• "The Industrial Economy Is Ending. What Comes Next?" - by Ted Shelton
• Thomas Piketty's book "Capital in the Twenty-First Century" - Wikipedia
• Nicholas Eberstadt's book "Men Without Work: America's Invisible Crisis" - Wikipedia
• Richard Sutton's essay "Bitter Lesson" - Wikipedia
• "Technofeudalism" - articles by Yanis Varoufakis

Music: Spike Protein, by Koi Discovery, available under CC0 1.0 Public Domain Declaration

Listen on: Apple Podcasts   Spotify

What happens if AI delivers major advances in capability and productivity, but also creates significant disruption to jobs, incomes, and public finances? That question sits at the heart of today’s episode.

Our guest is Adrian Brown, the Founder and Chief Executive of Windfall Trust, a nonprofit focused on helping governments and societies prepare for the economic consequences of advanced AI. Windfall describes itself not as a think tank, but as a “policy accelerator for the age of artificial intelligence”.

Their work starts from a simple premise: if AI systems significantly reshape the economy, then the question is not only how we build them, but how we prepare for their impacts, and how the gains are ultimately shared.

Before founding Windfall Trust, Adrian was the founding Executive Director of the Centre for Public Impact, worked as a policy advisor in the UK Cabinet Office, and held roles at McKinsey and the Boston Consulting Group.

Selected follow-ups:

• Windfall Trust
• Exercise Cygnus (Wikipedia)
• Erik Brynjolfsson (personal site)
• Canaries in the Coal Mine? Six Facts about the Recent Employment Effects of Artificial Intelligence (paper by Erik Brynjolfsson and colleagues)
• Anton Korinek (personal site)
• Windfall's UK Scenarios exercise
• The Economic Singularity (recent discussion paper by Calum Chace)
• Industrial Policy for the Intelligence Age (OpenAI)
• UK Government announcement of the formation of "The AI and Future of Work Unit"
• UK Chancellor Rachel Reeve's announcement of a new "AI Economics Institute"
• "AI will kill income tax" - episode of Robert Peston's podcast "The Rest is Money"
• Windfall Policy Atlas
• Task-Completion Time Horizons of Frontier AI Models (METR)

Music: Spike Protein, by Koi Discovery, available under CC0 1.0 Public Domain Declaration

Listen on: Apple Podcasts   Spotify

Could the future see the emergence and adoption of a new field of engineering called nucleonics, in which the energy of nuclear fusion is accessed at relatively low temperatures, producing abundant clean safe energy? This kind of idea has been discussed since 1989, when the claims of cold fusion first received media attention. It is often assumed that the field quickly reached a dead-end, and that the only scientists who continue to study it are cranks. However, as we’ll hear in this episode, there may be good reasons to keep an open mind about a number of anomalous but promising results.

Our guest is Jonah Messinger, who is a Winton Scholar and Ph.D. student at the Cavendish Laboratory of Physics at the University of Cambridge. Jonah is also a Research Affiliate at MIT, a Senior Energy Analyst at the Breakthrough Institute, and previously he was a Visiting Scientist and ThinkSwiss Scholar at ETH Zürich. His work has appeared in research journals, on the John Oliver show, and in publications of Columbia University. He earned his Master’s in Energy and Bachelor’s in Physics from the University of Illinois at Urbana-Champaign, where he was named to its Senior 100 Honorary.

Selected follow-ups:

• Jonah Messinger (The Breakthrough Institute)
• nucleonics.org
• U.S. Department of Energy Announces $10 Million in Funding to Projects Studying Low-Energy Nuclear Reactions (ARPA-E)
• How Anomalous Science Breaks Through - by Jonah Messinger
• Wolfgang Pauli (Wikiquote)
• Cold fusion: A case study for scientific behavior (Understanding Science)
• Calculated fusion rates in isotopic hydrogen molecules - by SE Koonin & M Nauenberg
• Known mechanisms that increase nuclear fusion rates in the solid state - by Florian Metzler et al
• Introduction to superradiance (Cold Fusion Blog)
• Peter L. Hagelstein - Professor at MIT
• Models for nuclear fusion in the solid state - by Peter Hagelstein et al
• Risk and Scientific Reputation: Lessons from Cold Fusion - by Huw Price
• Katalin Karikó (Wikipedia)
• “Abundance” and Its Insights for Policymakers - by Hadley Brown
• Identifying intellectual dark matter - by Florian Metzler and Jonah

Listen on: Apple Podcasts   Spotify

Our guest in this episode is Francesca Rossi. Francesca studied computer science at the University of Pisa in Italy, where she became a professor, before spending 20 years at the University of Padova. In 2015 she joined IBM's T.J. Watson Research Lab in New York, where she is now an IBM Fellow and also IBM's AI Ethics Global Leader.

Francesca is a member of numerous international bodies concerned with the beneficial use of AI, including being a board member at the Partnership on AI, a Steering Committee member and designated expert at the Global Partnership on AI, a member of the scientific advisory board of the Future of Life Institute, and Chair of the international conference on Artificial Intelligence, Ethics, and Society which is being held in Montreal in August this year.

From 2022 until 2024 she holds the prestigious role of the President of the AAAI, that is, the Association for the Advancement of Artificial Intelligence. The AAAI has recently held its annual conference, and in this episode, Francesca shares some reflections on what happened there.

Selected follow-ups:
https://researcher.watson.ibm.com/researcher/view.php?person=ibm-Francesca.Rossi2
https://en.wikipedia.org/wiki/Francesca_Rossi
https://partnershiponai.org/
https://gpai.ai/

Topics in this conversation include:

*) How a one-year sabbatical at the Harvard Radcliffe Institute changed the trajectory of Francesca's life
*) New generative AI systems such as ChatGPT expand previous issues involving bias, privacy, copyright, and content moderation - because they are trained on very large data sets that have not been curated
*) Large language models (LLMs) have been optimised, not for "factuality", but for creating language that is syntactically correct
*) Compared to previous AIs, the new systems impact a wider range of occupations, and they also have major implications for education
*) Are the "AI ethics" and "responsible AI" approaches that address the issues of existing AI systems also the best approaches for the "AI alignment" and "AI safety" issues raised by artificial general intelligence?
*) Different ideas on how future LLMs could acquire mastery, not only over language, but also over logic, inference, and reasoning
*) Options for combining classical AI techniques focussing on knowledge and reasoning, with the data-intensive approaches of LLMs
*) How "foundation models" allow training to be split into two phases, with a shorter supervised phase customising the output from a prior longer unsupervised phase
*) Even experts face the temptation to anthropomorphise the behaviour of LLMs
*) On the other hand, unexpected capabilities have emerged within LLMs
*) The interplay of "thinking fast" and "thinking slow" - adapting, for the context of AI, insights from Daniel Kahneman about human intelligence
*) Cross-fertilisation of ideas from different communities at the recent AAAI conference
*) An extension of that "bridge" theme to involve ideas from outside of AI itself, including the use of methods of physics to observe and interpret LLMs from the outside
*) Prospects for interpretability, explainability, and transparency of AI - and implications for trust and cooperation between humans and AIs
*) The roles played by different international bodies, such as PAI and GPAI
*) Pros and cons of including China in the initial phase of GPAI
*) Designing regulations to be future-proof, with parts that can change quickly
*) A

Listen on: Apple Podcasts   Spotify

In this episode, Tim Clement-Jones brings us up to date on the reactions by members of the UK's House of Commons to recent advances in the capabilities of AI systems, such as ChatGPT. He also looks ahead to larger changes, in the UK and elsewhere.

Lord Clement-Jones CBE, or Tim, as he prefers to be known, has been a very successful lawyer, holding senior positions at ITV and Kingfisher among others, and later becoming London Managing Partner of law firm DLA Piper.

He is better known as a politician. He became a life peer in 1998, and has been the Liberal Democrats’ spokesman on a wide range of issues. The reason we are delighted to have him as a guest on the podcast is that he was the chair of the AI Select Committee, Co-Chair of the All-Party Parliamentary Group on AI, and is now a member of a special inquiry on the use of AI in Weapons Systems.

Tim also has multiple connections with universities and charities in the UK.

Selected follow-up reading:
https://www.lordclementjones.org/
https://www.parallelparliament.co.uk/APPG/artificial-intelligence
https://arcs.qmul.ac.uk/governance/council/council-membership/timclement-jones.html

Topics in this conversation include:

*) Does "the Westminster bubble" understand the importance of AI?
*) Evidence that "the tide is turning" - MPs are demonstrating a spirit of inquiry
*) The example of Sir Peter Bottomley, the Father of the House (who has been an MP continuously since 1975)
*) New AI systems are showing characteristics that had not been expected to arrive for another 5 or 10 years, taking even AI experts by surprise
*) The AI duopoly (the US and China) and the possible influence of the UK and the EU
*) The forthcoming EU AI Act and the risk-based approach it embodies
*) The importance of regulatory systems being innovation-friendly
*) How might the EU support the development of some European AI tech giants?
*) The inevitability(?) of the UK needing to become "a rule taker"
*) Cynical and uncynical explanations for why major tech companies support EU AI regulation
*) The example of AI-powered facial recognition: benefits and risks
*) Is Brexit helping or hindering the UK's AI activities?
*) Complications with the funding of AI research in the UK's universities
*) The risks of a slow-down in the UK's AI start-up ecosystem
*) Looking further afield: AI ambitions in the UAE and Saudi Arabia
*) The particular risks of lethal autonomous weapons systems
*) Future conflicts between AI-controlled tanks and human-controlled tanks
*) Forecasts for the arrival of artificial general intelligence: 10-15 years from now?
*) Superintelligence may emerge from a combination of separate AI systems
*) The case for "technology-neutral" regulation

Music: Spike Protein, by Koi Discovery, available under CC0 1.0 Public Domain Declaration

Listen on: Apple Podcasts   Spotify

Advanced AI is currently pretty much a duopoly between the USA and China. The US is the clear leader, thanks largely to its tech giants – Google, Meta, Microsoft, Amazon, and Apple. China also has a fistful of tech giants – Baidu, Alibaba, and Tencent are the ones usually listed, but the Chinese government has also taken a strong interest in AI since Deep Mind’s Alpha Go system beat the world’s best Go player in 2016.

People in the West don’t know enough about China’s current and future role in AI. Some think its companies just copy their Western counterparts, while others think it is an implacable and increasingly dangerous enemy, run by a dictator who cares nothing for his people. Both those views are wrong.

One person who has been trying to provide a more accurate picture of China and AI in recent years is Jeff Ding, the author of the influential newsletter ChinAI.

Jeff grew up in Iowa City and is now an Assistant Professor of Political Science at George Washington University. He earned a PhD at Oxford University, where he was a Rhodes Scholar, and wrote his thesis on how past technological revolutions influenced the rise and fall of great powers, with implications for U.S.-China competition. After gaining his doctorate he worked at Oxford’s Future of Humanity Institute and Stanford’s Institute for Human-Centered Artificial Intelligence.

Selected follow-up reading:
https://jeffreyjding.github.io/
https://chinai.substack.com/
https://www.tortoisemedia.com/intelligence/global-ai/

Topics in this conversation include:

*) The Thucydides Trap: Is conflict inevitable as a rising geopolitical power approaches parity with an established power?
*) Different ways of trying to assess how China's AI industry compares with that of the U.S.
*) Measuring innovations in creating AI is different from measuring adoption of AI solutions across multiple industries
*) Comparisons of papers submitted to AI conferences such as NeurIPS, citations, patents granted, and the number of data scientists
*) The biggest misconceptions westerners have about China and AI
*) A way in which Europe could still be an important player alongside the duopoly
*) Attitudes in China toward data privacy and facial recognition
*) Government focus on AI can be counterproductive
*) Varieties of government industrial policy: the merits of encouraging decentralised innovation
*) The Titanic and the origin of Silicon Valley
*) Mariana Mazzucato's question: "Who created the iPhone?"
*) Learning from the failure of Japan's 5th Generation Computers initiative
*) The evolution of China's Social Credit systems
*) Research by Shazeda Ahmed and Jeremy Daum
*) Factors encouraging and discouraging the "splinternet" separation of US and Chinese tech ecosystems
*) Connections that typically happen outside of the public eye
*) Financial interdependencies
*) Changing Chinese government attitudes toward Chinese Internet giants
*) A broader tension faced by the Chinese government
*) Future scenarios: potential good and bad developments
*) Transnational projects to prevent accidents or unauthorised use of powerful AI systems

Music: Spike Protein, by Koi Discovery, available under CC0 1.0 Public Domain Declaration

Listen on: Apple Podcasts   Spotify

Peter James is one of the world’s most successful crime writers. His "Roy Grace" series, about a detective in Brighton, England, near where Peter lives, has produced a remarkable 19 consecutive Sunday Times Number One bestsellers. His legions of devoted fans await each new release eagerly. The books have been televised, with the third series of "Grace", starting John Simm, being commissioned for next year.

Peter has worked in other genres too, having written 36 novels altogether. When Calum first met Peter in the mid-1990s, Peter's science fiction novel “Host” was generating rave reviews. It was the world’s first electronically published novel, and a copy of its floppy disc version is on display in London’s Science Museum.

Peter is also a self-confessed petrol-head, with an enviable collection of classic cars, and a pretty successful track record of racing some of them. The discussion later in the episode addresses the likely arrival of self-driving cars. But we start with the possibility of mind uploading, which is the subject of “Host”.

Selected follow-up reading:
https://www.peterjames.com/
https://www.alcor.org/

Topics in this conversation include:

*) Peter's passion for the future
*) The transformative effect of the 1990 book "Great Mambo Chicken and the Transhuman Condition"
*) A Christmas sojourn at MIT and encounters with AI pioneer Marvin Minsky
*) The origins of the ideas behind "Host"
*) Meeting Alcor, the cryonics organisation, in Riverside California
*) How cryonics has evolved over the decades
*) "The first person to live to 200 has already been born"
*) Quick summaries of previous London Futurists Podcast episodes featuring Aubrey de Grey and Andrew Steele
*) The case for doing better than nature
*) Peter's novel "Perfect People" and the theme of "designer babies"
*) Possible improvements in the human condition from genetic editing
*) The risk of a future "genetic underclass"
*) Technology divides often don't last: consider the "fridge divide" and the "smartphone divide"
*) Calum's novel "Pandora's Brain"
*) Why Peter is comfortable with the label "transhumanist"
*) Various ways of reading (many) more books
*) A thought experiment involving a healthy 99 year old
*) If people lived a lot longer, we might take better care of our planet
*) Peter's views on technology assisting writers
*) Strengths and weaknesses of present-day ChatGPT as a writer
*) Prospects for transhumans to explore space
*) The "bunker experiments" into the circadian cycle, which suggest that humans naturally revert to a daily cycle closer to 26 hours than 24 hours
*) Possible answers to Fermi's question about lack of any sign of alien civilisations
*) Reflections on "The Pale Blue Dot of Earth" (originally by Carl Sagan)
*) The likelihood of incredible surprises in the next few decades
*) Pros and cons of humans driving on public roads (especially when drivers are using mobile phones)
*) Legal and ethical issues arising from autonomous cars
*) Exponential change often involves a frustrating slow phase before fast breakthroughs
*) Anticipating the experience of driving inside immersive virtual reality
*) The tragic background to Peter's book "Possession"
*) A concluding message from the science fiction writer Kurt Vonnegut

Music: Spike Protein, by Koi Discovery, available under CC0 1.0 Public Domain Declaration

Listen on: Apple Podcasts   Spotify

Our guest in this episode is a Briton who is based in Berlin, namely Andrew Steele. Earlier in his life Andrew spent nine years at the University of Oxford where, among other accomplishments, he gained a PhD in physics. His focus switched to computational biology, and he held positions at Cancer Research UK and the Francis Crick Institute.

Along the way, Andrew decided that aging was the single most important scientific challenge of our time. This led him to write the book "Ageless: The New Science of Getting Older Without Getting Old". There are a lot of books these days about the science of slowing, stopping, and even reversing aging, but Andrew's book is perhaps the best general scientific introduction to this whole field.

Selected follow-ups:
https://andrewsteele.co.uk/
https://www.youtube.com/DrAndrewSteele
https://ageless.link/

Topics in this conversation include:
*) The background that led Andrew to write his book "Ageless"
*) A graph that changed a career
*) The chance of someone dying in the next year doubles every eight years they live
*) For tens of thousand of years, human life expectancy didn't change 
*) In recent centuries, the background mortality rate has significantly decreased, but the eight year "Gompertz curve" doubling of mortality remains unchanged
*) Some animals do not have this mortality doubling characteristic; they are said to be "negligibly senescent", "biologically immortal", or "ageless"
*) An example: Galapagos tortoises
*) The concept of "hallmarks of aging" - and different lists of these hallmarks
*) Theories of aging: wear-and-tear vs. programmed obsolescence
*) Evolution and aging: two different strategies that species can adopt
*) Wear-and-tear of teeth - as seen from a programmed aging point-of-view
*) The case for a pragmatic approach
*) Dietary restriction and healthier aging
*) The potential of computational biology system models to generate better understanding of linkages between different hallmarks of aging
*) Might some hallmarks, for example telomere shortening or epigenetic damage, prove more fundamental than others?
*) Special challenges posed by damage in the proteins in the scaffolding between cells
*) What's required to accelerate the advent of "longevity escape velocity"
*) Excitement and questions over the funding available to Altos Labs
*) Measuring timescales in research dollars rather than years
*) Reasons for optimism for treatments of some of the hallmarks, for example with senolytics, but others aren't being properly addressed
*) Breakthrough progress with the remaining hallmarks could be achieved with $5-10B investment each
*) Adding some extra for potential unforeseen hallmarks, that sums to a total of around $100B before therapies for all aspects of aging could be in major clinical trials
*) Why such an expenditure is in principle relatively easily affordable
*) Reflections on moral and ethical objections to treatments against aging
*) Overpopulation, environmental strains, resource sustainability, and net zero impact
*) Aging as the single largest cause of death in the world - in all countries
*) Andrew's current and forthcoming projects, including a book on options for funding science with the biggest impact
*) Looking forward to "being more tortoise".

Music: Spike Protein, by Koi Discovery, available under CC0 1.0 Public Domain Declaration

Listen on: Apple Podcasts   Spotify

It is nearly 40 years since our guest in this episode, pioneering transhumanist Natasha Vita-More, created the first version of the Transhumanist Manifesto. Since that time, Natasha has established numerous core perspectives, values, and actions in the global transhumanist family.

Natasha joins us in this episode to share her observations on how transhumanism has evolved over the decades, and to reflect on her work in building the movement—from practice-based approaches, scientific contributions, and theoretical innovations.

Areas we explore include: How has Natasha's work seeded the global growth of transhumanism? What are the main advances over the years that she particularly values? And what are the disappointments?

We also look to the future: What are her hopes and expectations for the next ten years of transhumanism?

Music: Spike Protein, by Koi Discovery, available under CC0 1.0 Public Domain Declaration

Selected follow-up reading:

https://natashavita-more.com/
https://www.fightaging.org/archives/2004/02/vital-progress-summit/
http://www.extropy.org/proactionaryprinciple.htm
https://metanexus.net/transhumanism-and-its-critics/
https://whatistranshumanism.org/
https://www.alcor.org/library/persistence-of-long-term-memory-in-vitrified-and-revived-simple-animals/
https://waitbutwhy.com/2016/03/cryonics.html
F. M. Esfandiary: https://archives.nypl.org/mss/4846
https://www.maxmore.com/
The World’s Most Dangerous Idea? https://nickbostrom.com/papers/dangerous
https://theconversation.com/the-end-of-history-francis-fukuyamas-controversial-idea-explained-193225
https://www.humanityplus.org/
https://transhumanist-studies.teachable.com/
Anyone Can Code, Ethiopia: https://icogacc.com/
https://afrolongevity.taffds.org/

Listen on: Apple Podcasts   Spotify

Our guest in this episode is the scientist and science fiction author Davin Brin, whose writings have won the Hugo, Locus, Campbell, and Nebula Awards. His style is sometimes called 'hard science fiction'. This means his narratives feature scientific or technological change that is plausible rather than purely magical. The scenarios he creates are thought-provoking as well as entertaining. His writing inspires readers but also challenges them, with important questions not just about the future, but also about the present.

Perhaps his most famous non-fiction work is his book "The Transparent Society: Will Technology Force Us to Choose Between Privacy and Freedom?", first published in 1998. With each passing year it seems that the questions and solutions raised in that book are becoming ever more pressing. One aspect of this has been called Brin's Corollary to Moore's Law: Every year, the cameras will get smaller, cheaper, more numerous and more mobile.

David also frequently writes online about topics such as space exploration, attempts to contact aliens, homeland security, the influence of science fiction on society and culture, the future of democracy, and much more besides.

Topics discussed in this conversation include:

*) Reactions to reports of flying saucers
*) Why photographs of UFOs remain blurry
*) Similarities between reports of UFOs and, in prior times, reports of elves
*) Replicating UFO phenomena with cat lasers
*) Changes in attitudes by senior members of the US military
*) Appraisals of the Mars Rovers
*) Pros and cons of additional human visits to the moon
*) Why alien probes might be monitoring this solar system from the asteroid belt
*) Investigations of "moonlets" in Earth orbit
*) Looking for pi in the sky
*) Reasons why life might be widespread in the galaxy - but why life intelligent enough to launch spacecraft may be rare
*) Varieties of animal intelligence: How special are humans?
*) Humans vs. Neanderthals: rounds one and two
*) The challenges of writing about a world that includes superintelligence
*) Kurzweil-style hybridisation and Mormon theology
*) Who should we admire most: lone heroes or citizens?
*) Benefits of reciprocal accountability and mutual monitoring (sousveillance)
*) Human nature: Delusions, charlatans, and incantations
*) The great catechism of science
*) Two levels at which the ideas of a transparent society can operate
*) "Asimov's Laws of Robotics won't work"
*) How AIs might be kept in check by other AIs
*) The importance of presenting gedanken experiments

Fiction mentioned (written by David Brin unless noted otherwise):
The Three-Body Problem (Liu Cixin)
Existence
The Sentinel (Arthur C. Clarke)
Startide Rising
The Uplift War
Kiln People
The Culture Series (Iain M. Banks)
The Expanse (James S.A. Corey)
The Postman (the book and the film)
Stones of Significance
Fahrenheit 451 (Ray Bradbury)

Music: Spike Protein, by Koi Discovery, available under CC0 1.0 Public Domain Declaration

Selected follow-up reading:
http://www.davidbrin.com/
http://davidbrin.blogspot.com/2021/07/whats-really-up-with-uaps-ufos.html

Listen on: Apple Podcasts   Spotify