Can we build AI without losing control over it? – Sam Harris

San Harris (author of The Moral Landscape and host of the Waking Up podcast) discusses the need for AI Safety – while fun to think about, we are unable to “martial an appropriate emotional response” to improvements in AI and automation and the prospect of dangerous AI – it’s a failure of intuition to respond to it like one would a sci-fi like doom scenario.

Scared of superintelligent AI? You should be, says neuroscientist and philosopher Sam Harris — and not just in some theoretical way. We’re going to build superhuman machines, says Harris, but we haven’t yet grappled with the problems associated with creating something that may treat us the way we treat ants.

Anders Sandberg -The Technological Singularity

Anders Sandberg.00_23_53_16.Still031Anders gives a short tutorial on the Singularity – clearing up confusion and highlighting important aspects of the Technological Singularity and related ideas, such as accelerating change, horizons of predictability, self-improving artificial intelligence, and the intelligence explosion.

Tutorial Video:

Points covered in the tutorial:

  • The Mathematical Singularity
  • The Technological Singularity: A Horizon of predictability
  • Confusion Around The Technological Singularity
  • Drivers of Accelerated Growth
  • Technology Feedback Loops
  • A History of Coordination
  • Technological Inflection Points
  • Difficult of seeing what happens after an Inflection Point
  • The Intelligence Explosion
  • An Optimisation Power Applied To Itself
  • Group Minds
  • The HIVE Singularity: A Networked Global Mind
  • The Biointelligence explosion
  • Humans are difficult to optimise

An Overview of Models of the Technological Singularity

anders-sandberg-technology-feedback-loopsSee Anders’ paper ‘An overview of models of technological singularity
This paper reviews different definitions and models of technological singularity. The models range from conceptual sketches to detailed endogenous growth models, as well as attempts to fit empirical data to quantitative models. Such models are useful for examining the dynamics of the world-system and possible types of future crisis points where fundamental transitions are likely to occur. Current models suggest that, generically, even small increasing returns tends to produce radical growth. If mental capital becomes copyable (such as would be the case for AI or brain emulation) extremely rapid growth would also become likely.

[The] Technological singularity is of increasing interest among futurists both as a predicted possibility in the midterm future and as subject for methodological debate. The concept is used in a variety of contexts, and has acquired an unfortunately large number of meanings. Some versions stress the role of artificial intelligence, others refer to more general technological change. These multiple meanings can overlap, and many writers use combinations of meanings: even Vernor Vinge’s seminal essay that coined the term uses several meanings. Some of these meanings may imply each other but often there is a conflation of different elements that likely (but not necessarily) occur in parallel. This causes confusion and misunderstanding to the extent that some critics argue that the term should be avoided altogether. At the very least the term ‘singularity’ has led to many unfortunate assumptions that technological singularity involves some form of mathematical singularity and can hence be ignored as unphysical.Anders Sandberg

A list of models described in the paper:

A. Accelerating change

Exponential or superexponential technological growth (with linked economical growth and social change) (Ray Kurzweil (Kur05), John Smart (Smang))

B. Self improving technology

Better technology allows faster development of new and better technology. (Flake (Fla06))

C. Intelligence explosion

Smarter systems can improve themselves, producing even more intelligence in a strong feedback loop. (I.J. Good (Goo65), Eliezer Yudkowsky)

D. Emergence of superintelligence

(Singularity Institute) 1

E. Prediction horizon

Rapid change or the emergence of superhuman intelligence makes the future impossible to predict from our current limited knowledge and experience. (Vinge, (Vin93))

F. Phase transition

The singularity represents a shift to new forms of organisation. This could be a fundamental difference in kind such as humanity being succeeded by posthuman or artificial intelligences,
a punctuated equilibrium transition or the emergence of a new meta-system level. (Teilhard de Chardin, Valentin Turchin (Tur77), Heylighen (Hey07))

G. Complexity disaster

Increasing complexity and interconnectedness causes increasing payoffs, but increases instability. Eventually this produces a crisis, beyond which point the dynamics must be different.
(Sornette (JS01), West (BLH+07))

H. Inflexion point

Large-scale growth of technology or economy follows a logistic growth curve. The singularity represents the inflexion point where change shifts from acceleration to de-acceleration. (Extropian
FAQ, T. Modis (Mod02))

I. Infinite progress

The rate of progress in some domain goes to infinity in nite time. (Few, if any, hold this to be plausible 2 )


Many thanks for watching!

Consider supporting SciFuture by:
a) Subscribing to the YouTube channel:
b) Donating via Patreon: and/or
c) Sharing the media SciFuture creates

Science, Technology & the Future:

AI: The Story So Far – Stuart Russell

stuart russell - redAwesome to have Stuart Russell discussing AI Safety – a very important topic. Too long have people been associating the idea of AI safety issues with Terminator – unfortunately the human condition seems such that people often don’t give themselves permission to take seriously non-mainstream ideas unless they see a tip of the hat from an authority figure.

During the presentation Stuart brings up a nice quote by Norbert Wiener:

If we use, to achieve our purposes, a mechanical agency with whose operation we cannot efficiently interfere once we have started it, because the action is so fast and irrevocable that we have not the data to intervene before the action is complete, then we had better be quite sure that the purpose put into the machine is the purpose which we really desire and not merely a colorful imitation of it.Norbert Wiener

P.s. Stuart Russell co-authored AI A Modern Approach with Peter Norvig – arguably the most popular textbook on AI theory.

The lecture was presented at the 2016 Colloquium Series on Robust and Beneficial AI (CSRBAI) hosted by the Machine Intelligence Research Institute (MIRI) and Oxford’s Future of Humanity Institute (FHI).

What I’m finding is that senior people in the field who have never publicly evinced any concern before are privately thinking that we do need to take this issue very seriously, and the sooner we take it seriously the better.Stuart Russell

Video of presentation:


The field [of AI] has operated for over 50 years on one simple assumption: the more intelligent, the better. To this must be conjoined an overriding concern for the benefit of humanity. The argument is very simple:

1. AI is likely to succeed.
2. Unconstrained success brings huge risks and huge benefits.
3. What can we do now to improve the chances of reaping the benefits and avoiding the risks?

Some organizations are already considering these questions, including the Future of Humanity Institute at Oxford, the Centre for the Study of Existential Risk at Cambridge, the Machine Intelligence Research Institute in Berkeley, and the Future of Life Institute at Harvard/MIT. I serve on the Advisory Boards of CSER and FLI.

Just as nuclear fusion researchers consider the problem of containment of fusion reactions as one of the primary problems of their field, it seems inevitable that issues of control and safety will become central to AI as the field matures. The research questions are beginning to be formulated and range from highly technical (foundational issues of rationality and utility, provable properties of agents, etc.) to broadly philosophical.

– Stuart Russell (Quote Source)

Nick Bostrom: Why Focus on Existential Risk related to Machine Intelligence?

One can think of Existential Risk as a subcategory of a Global Catastrophic Risk – while GCR’s are really bad, civilization has the potential to recover from such a global catastrophic disaster.
An existential Risk is one in which there is no chance of recoverability. An example of the sort of disaster that fits the category of existential risk is human extinction which reduces the probability of [human] lives worth living – theories of value that imply even relatively small reductions in net existential risk have enormous expected value mostly fall under population ethics that consider an average or total utilitarian view of the well-being of the future of life in the universe.  Since we haven’t seen any convincing evidence of life outside earth’s gravity well, it may be that there is no advanced technologically capable life elsewhere in the observable universe.  If we value lives worth living, and lots of lives worth living, we might also value filling the uninhabited parts of the universe with lives worth living – and arguably we need an advanced technologically able civilization to achieve this.  Hence, if humans become extinct it may be that evolution will never again produce a life form capable of escaping the gravity well and colonizing the universe with valuable life.

Here we focus on the reasons to focus on Existential Risk related to machine intelligence.

Say machine intelligence is created with a theory of value outside of, contradictory to, or simply different enough to one in which valued human existence, or the existence of valuable life in the universe.  Also imagine that this machine intelligence could enact on it’s values in an exacting manner – it may cause humanity to become extinct on purpose, or as a side effect of implementing it’s values.

The paper ‘Existential Risk Prevention as Global Priority‘ by Nick Bostrom clarifies the concept of existential risk further:

Existential risks are those that threaten the entire future of humanity. Many theories of value imply that even relatively small reductions in net existential risk have enormous expected value. Despite their importance, issues surrounding human-extinction risks and related hazards remain poorly understood. In this paper, I clarify the concept of existential risk and develop an improved classification scheme. I discuss the relation between existential risks and basic issues in axiology, and show how existential risk reduction (via the maxipok rule) can serve as a strongly action-guiding principle for utilitarian concerns. I also show how the notion of existential risk suggests a new way of thinking about the ideal of sustainability.

Interview with Nick Bostrom on Machine Intelligence and XRisk

I had the pleasure of doing an interview with Oxford philosopher Nick Bostrom on XRisk:

Transcription of interview:

In recent couple of years we’ve been focusing quite heavily on machine intelligence partly because it seems to raise some significant existentialist down the road part also because relatively little attention has been given to this risk. So when we are prioritizing what we want to spend our time researching then one variable that we take into account is how important is this topic that we could research? But another is how many other people are there who are already studying it? Because the more people who already studying it – the smaller the difference that having a few extra minds focusing on that topic.
So, say the topic of peace and war and how you can try to avoid international conflict is a very important topic – and many existential risks will be reduced if there is more global corporation.
However it is also hard to see how a very small group of people could make a substantial difference to today’s risk of arms races and wars. There is a big interest involved in this and so many people already working either on disarmament and peace and/or military strength that it’s an area where it would be great to make a change – but it’s hard to make a change if there are a smaller number people by contrast with something like the risk from machine intelligence and the risk of Super-Intelligence.
Only been a relatively small number of people have been thinking about this and there might be some low-hanging fruit there – some insights that might make a big difference. So that’s one of the criteria.
Now we are also looking at other existential risks and we are also looking at things other than existential risk like – with try to get a better understanding of what humanity’s situation is in the world and so we have been thinking some about the Fermi Paradox for example, some methodological tools that you need like observation selection theory how you can reason about these things. And to some extent also more near term impacts of technology and of course the opportunities involved in all of this – is that always worth to remind oneself that although enormous technological powers will pose great new dangers including existential risks they also of course make it possible to achieve enormous amount of good.
So one should bear in mind this ..the opportunities as well that are unleashed with technological advance.

About Professor Nick Bostrom

Director & James Martin Research Fellow

Bostrom Xrisk 2Nick Bostrom is Professor in the Faculty of Philosophy at Oxford University and founding Director of the Future of Humanity Institute and of the Programme on the Impacts of Future Technology within the Oxford Martin School. He is the author of some 200 publications, including Anthropic Bias (Routledge, 2002), Global Catastrophic Risks (ed., OUP, 2008), and Human Enhancement (ed., OUP, 2009), and a forthcoming book on Superintelligence. He previously taught at Yale, and he was a Postdoctoral Fellow of the British Academy. Bostrom has a background in physics, computational neuroscience, and mathematical logic as well as philosophy.

He is best known for his work in five areas: (i) the concept of existential risk; (ii) the simulation argument; (iii) anthropics (developing the first mathematically explicit theory of observation selection effects); (iv) transhumanism, including related issues in bioethics and on consequences of future technologies; and (v) foundations and practical implications of consequentialism. He is currently working on a book on the possibility of an intelligence explosion and on the existential risks and strategic issues related to the prospect of machine superintelligence.

In 2009, he was awarded the Eugene R. Gannon Award (one person selected annually worldwide from the fields of philosophy, mathematics, the arts and other humanities, and the natural sciences). He has been listed in the FP 100 Global Thinkers list, the Foreign Policy Magazine’s list of the world’s top 100 minds. His writings have been translated into more than 21 languages, and there have been some 80 translations or reprints of his works. He has done more than 500 interviews for TV, film, radio, and print media, and he has addressed academic and popular audiences around the world.


Personal Web:

FHI Bio:

Also consider joining the Facebook Group on Existential Risk:

Conference: Thinking Machines in the Physical World

“Thinking Machines in the Physical World” invites cross-disciplinary conversations about the opportunities and threats presented by advances in cognitive computing:
  – What concrete, real-world possibilities does intelligence-focused technology open up?
  – What potential effects will “smart computers” exert on labor and jobs around the globe?
  – What are the broader social implications of these changes?

When: Wednesday, July 13, 2016 8:30 AM until Friday ~6pm (then dinner)
Where: Melbourne Uni Law School Building, Level 10 185 Pelham Street, Carlton

Keynotes (see details here):

Prof Brian Anderson – Distinguished Professor at ANU College of Engineering and Computer Science.

Dr James Hughes – Executive Director of the Institute for Ethics and Emerging Technologies.

Prof M. Vidyasagar – Cecil & Ida Green Chair in Systems Biology Science

Prof Judy Wajcman – Anthony Giddens Professor of Sociology, London School of Economics

Dr. Juerg von Kaenel, IBM Research – Cognitive Computing – IBM Watson

Register here | Main website | Program

Professor Graeme Clark, AC Laureate Professor Emeritus  says “It gives me great pleasure to have the opportunity to welcome your interest in the work of Norbert Wiener and invite you to Melbourne to participate in this important conference.”

Official Website:
Facebook Event:

Peter Singer & David Pearce on Utilitarianism, Bliss & Suffering

Moral philosophers Peter Singer & David Pearce discuss some of the long term issues with various forms of utilitarianism, the future of predation and utilitronium shockwaves.

Topics Covered

Peter Singer

– long term impacts of various forms of utilitarianism
– Consciousness
– Artificial Intelligence
– Reducing suffering in the long run and in the short term
– Practical ethics
– Pre-implantation genetic screening to reduce disease and low mood
– Lives today are worth the same as lives in the future – though uncertainty must be brought to bear in deciding how one weighs up the importance of life
– The Hedonistic Imperative and how people react to it
– Correlation to high hedonic set points with productivity
existential risks and global catastrophic risks
– Closing factory farms

David Pearce

– Veganism and reducitarianism
– Red meat vs white meat – many more chickens are killed per ton of meat than beef
– Valence research
– Should one eliminate the suffering? And should we eliminate emotions of happiness?
– How can we answer the question of how far suffering is present in different life forms (like insects)?

Talk of moral progress can make one sound naive. But even the darkest cynic should salute the extraordinary work of Peter Singer to promote the interests of all sentient beings.David Pearce


Many thanks for watching!
– Support me via Patreon:
– Please Subscribe to this Channel:…
– Science, Technology & the Future website:

Is there a Meaningful Future for Non-Optimal Moral Agents?

In an interview last year, I had a discussion with John Danaher on the Hedonistic Imperative & Superintelligence – a concern he has with HI is that it denies or de-emphasises some kind of moral agency – in moral theory there is a distinction between moral agents (being a responsible actor able to make moral decisions, influence direction of moral progress, shapes its future, and owes duties to others) and moral patients who may be deemed to have limited or no grounds for moral agency/autonomy/responsibility – they are simply a recipient of moral benefits – in contrast to humans, animals could be classified as moral patients – (see Stanford writing on Grounds for Moral Status).

As time goes on, the notion of strong artificial intelligence leading to Superintelligence (which may herald in something like an Intelligence Explosion) and ideas like the hedonistic imperative becomes less sensational sci-fi concepts and more like visions of realizable eventualities. Thinking about moral endpoints comes to me a paradoxical feeling of triumph and disempowerment.

John’s concern is that ensuring the well-being of humans (conscious entities) is consistent with denying their moral agency – minimizing their capacity to act – that there is a danger that the outcome of HI or an Intelligence Explosion may result in sentient life being made very happy forever, but unable to make choices – with a focus on a future entirely based on bliss whilst ignoring other aspects of what makes for a valuable or worthwhile existence.

Artificial Heart chipsSo even if we have a future where a) we are made very happy and b) we are subject to a wide variety of novelty (which I argue for in Novelty Utilitarianism) without some kind of self-determination we may not be able to enjoy part of what arguably makes for a worthwhile existence.

If the argument for moral agency is completely toppled by the argument against free will then I can see why there would be no reason for it – and that bliss/novelty may be enough – though I personally haven’t been convinced that this is the case.

Also the idea that moral agency and novelty should be ranked as auxiliary aspects to the main imperative of reducing suffering/increasing bliss seems problematic – I get the sense that they (agency/novelty) could easily be swapped out for most non-optimal moral agents in the quest for -suffering/+bliss troublesome.
The idea that upon evaluating grounds for moral status, our ethical/moral quotient may not match or even come close to a potential ethical force of a superintelligence is also troubling. If we are serious about the best ethical outcomes, when the time comes, should we be committed to resigning all moral agency to agents that are more adept at producing peek moral outcomes?
ancillary-one-esk-glitchIs it really possible for non-optimal agents to have a meaningful moral input in a universe where they’ve been completely outperformed by moral machines? Is a life of novelty & bliss the most optimal outcome we can hope for?

There probably should be some more discussion on trade-offs between moral agency, peek experience and novelty.

Discussion in this video here starts at 24:02

Below is the whole interview with John Danaher:

The long-term future of AI (and what we can do about it) : Daniel Dewey at TEDxVienna

daniel deweyThis has been one of my favourite simple talks on AI Impacts – Simple, clear and straight to the point. Recommended as an introduction to the ideas (referred to in the title).

I couldn’t find the audio of this talk at TED – it has been added to


Daniel Dewey is a research fellow in the Oxford Martin Programme on the Impacts of Future Technology at the Future of Humanity Institute, University of Oxford. His research includes paths and timelines to machine superintelligence, the possibility of intelligence explosion, and the strategic and technical challenges arising from these possibilities. Previously, Daniel worked as a software engineer at Google, did research at Intel Research Pittsburgh, and studied computer science and philosophy at Carnegie Mellon University. He is also a research associate at the Machine Intelligence Research Institute.


Brian Greene on Artificial Intelligence, the Importance of Fundamental Physics, Alien Life, and the Possible Future of Our Civilization

March 14th was Albert Einstein’s birthday, and also PI day, so it was a fitting day to be interviewing well known theoretical physicist and string theorist Brian Greene – the author of a number of books including, The Elegant Universe, Icarus at the Edge of Time, The Fabric of the Cosmos, and The Hidden Reality!
Think-Inc-logo2Many thanks to Suzi and Desh at THINKINC for helping organize this interview & for bringing Brian Greene to Australia for a number of shows (March 16 in Perth, March 18 in Sydney and March 19 in Melbourne) – check out for more info!

Audio recording of the interview:

About the Interview with Brian Greene

Brian Greene discusses implications Artificial Intelligence and news of DeepMind AI (AlphaGo) beating the world grand champion in the board game Go.  He then discusses physics string theory, the territory of opinion on grand unifying theories of physics, the importance of supporting fundamental science, the possibility of alien life, the possible future of our space-faring civilization and of course gravitational waves!

In answer to the question on the importance of supporting fundamental research in science, Brain Greene said:

I tell them to wake up! Wake up and recognize that fundamental science has radically changed the way they live their lives today. If any of these individuals have a cell phone, or a personal computer, or perhaps they themselves or loved ones has been saved by an MRI machine.. I mean any of these devices rely on integrated circuits, which they themselves rely on quantum physics – so IF those folks who were in charge in the 1920s had have said, ‘hey you guys working on quantum physics, that doesn’t seem to be relevant to anything in the world around as so were going to cut your funding – well those people would have short circuited on of the greatest revolutions that our species has gone through – the information age, the technological age – so the bottom line is we need to support fundamental research because we know historically that when you gain a deep understanding of how things work – we can often leverage that to then manipulate the world around us in spectacular ways! And that needs to be where our fundamental focus remains – in science!


Layered art of Brian Greene, background and series titleBrian Randolph Greene is an American theoretical physicist and string theorist. He has been a professor at Columbia University since 1996 and chairman of the World Science Festival since co-founding it in 2008. Greene has worked on mirror symmetry, relating two different Calabi–Yau manifolds (concretely, relating the conifold to one of its orbifolds). He also described the flop transition, a mild form of topology change, showing that topology in string theory can change at the conifold point.

Greene has become known to a wider audience through his books for the general public, The Elegant Universe, Icarus at the Edge of Time, The Fabric of the Cosmos, The Hidden Reality, and related PBS television specials. He also appeared on The Big Bang Theory episode “The Herb Garden Germination“, as well as the films Frequency and The Last Mimzy. He is currently a member of the Board of Sponsors of the Bulletin of the Atomic Scientists.


Many thanks for listening!
Support me via Patreon
Please Subscribe to the YouTube Channel
Science, Technology & the Future on the web


Can Intelligence Explode? – Marcus Hutter at Singularity Summit Australia 2012

Abstract: The technological singularity refers to a hypothetical scenario in which technological advances virtually explode. The most popular scenario is the creation of super-intelligent algorithms that recursively create ever higher intelligences. After a short introduction to this intriguing potential future, I will elaborate on what it could mean for intelligence to explode. In this course, I will (have to) provide a more careful treatment of what intelligence actually is, separate speed from intelligence explosion, compare what super-intelligent participants and classical human observers might experience and do, discuss immediate implications for the diversity and value of life, consider possible bounds on intelligence, and contemplate intelligences right at the singularity.



Slides (pdf):
Slides (PowerPoint):
Paper: M.Hutter, Can Intelligence Explode, Journal of Consciousness Studies, Vol.19, Nr 1-2 (2012) pages 143–166.

Also see:

Marcus Hutter (born 1967) is a German computer scientist and professor at the Australian National University. Hutter was born and educated in Munich, where he studied physics and computer science at the Technical University of Munich. In 2000 he joined Jürgen Schmidhuber’s group at the Swiss Artificial Intelligence lab IDSIA, where he developed the first mathematical theory of optimal Universal Artificial Intelligence, based on Kolmogorov complexity and Ray Solomonoff’s theory of universal inductive inference. In 2006 he also accepted a professorship at the Australian National University in Canberra.

Hutter’s notion of universal AI describes the optimal strategy of an agent that wants to maximize its future expected reward in some unknown dynamic environment, up to some fixed future horizon. This is the general reinforcement learning problem. Solomonoff/Hutter’s only assumption is that the reactions of the environment in response to the agent’s actions follow some unknown but computable probability distribution.


Professor Marcus Hutter

Research interests:

Artificial intelligence, Bayesian statistics, theoretical computer science, machine learning, sequential decision theory, universal forecasting, algorithmic information theory, adaptive control, MDL, image processing, particle physics, philosophy of science.


Marcus Hutter is Professor in the RSCS at the Australian National University in Canberra, Australia. He received his PhD and BSc in physics from the LMU in Munich and a Habilitation, MSc, and BSc in informatics from the TU Munich. Since 2000, his research at IDSIA and now ANU is centered around the information-theoretic foundations of inductive reasoning and reinforcement learning, which has resulted in 100+ publications and several awards. His book “Universal Artificial Intelligence” (Springer, EATCS, 2005) develops the first sound and complete theory of AI. He also runs the Human Knowledge Compression Contest (50’000€ H-prize).