Artificial Intelligence Will Erase Humanity from the Earth by 2100
I will argue in this essay that humanity today faces two serious threats from artificial intelligence (AI). The first threat is financial, a threat to our livelihoods; the second threat is existential, a threat to our existence, to the survival of our race.
I will argue that we need to start on mitigation courses of action now. These mitigations will challenge some of our dearly held views, from how to ensure the livelihoods of our families, to what it means to be human. If considered now, debated, and acted on together, these actions will hopefully ensure our survival and freedom on this planet well beyond 2100. Lastly, I will issue you a call to action.
The risks that AI poses are significant. The seriousness of these risks can be discerned from the rapid, and accelerating rate of developments in AI, robotics, machine learning, AI weapons systems and related fields of technology. The rising level of AI risks can be validated with the evidence over recent years of broadening and deepening impacts, as these advances are released into and disperse through our societies and economies. One can argue about the likelihood of realization of these risks. One cannot argue that these risks are not significant. From a perspective of risk management, and with a historical view that we human explore the full spectrum of capabilities of new technologies, we must mitigate against these risks now.
We cannot afford to take a Pollyanna like view that AI is only about human enhancement and empowerment, and that the huge, generational tsunami of job losses, of which the first waves are already pummeling us are just a ‘structural adjustment,’ or ‘creative destruction’ as some neo-Schumpterian economists might say.
At this point, if you’re still reading, a short interlude. You may want to know who am I to make such a provocative, even alarmist prognosis? Am I promoting my latest book? No .. though that’s an idea with merit. Am I a zealot proselytizing for a religion with end of world views ? No. Do I with some regularity come out with apocalyptic statements? No.
I have a commerce/ IT background, an applied interest in AI, machine learning, a reasonable ability for pattern matching and inference, and a healthy respect for history, gained from yet inadequate reading on the capacity for humanity to repeatedly explore the full spectrum of potentials for new technologies. I have a family. I am optimistic for our future. For my children, I see great potentials ahead, but also great dangers over the next 80 years, to 2100. Thus, it imperative for me, for us to take action, to not sit back and ignore, or hope that someone else will do the hard work. We can see the results of such attitudes playing out now on another stage.
I’m considering writing a novel on how this collision course we’re heading for may play out. I’ve commented on Twitter in pithily brief prose on a semi-regular flow of articles; ‘experts’ telling us not to be concerned, that all will be fine with AI. I finally decided that it was time to write something longer in response than 140 characters on Feb 23. On LinkedIn Pulse I read an article, titled ‘Why Automation Will Create, Not Destroy, Jobs’ (5). I didn’t find support in the article for the prognosis that automation will net create, not destroy jobs. I added a comment to this effect.
Mention of the Industrial Revolution (1760–1840) does not provide an automatic Q.E.D. (like an all powerful, winning argument) on impacts of AI for article authors. The Industrial Revolution involved human augmentation. This coming revolution involves human replacement. This is a fundamental difference. Blue collar manufacturing jobs are being replaced at a rising rate by the hundreds of thousands of increasing sophisticated robots deployed yearly into factories and workplaces across the world. In transportation, taxi and bus drivers are starting to be replaced by autonomous vehicles; next will be the long haul carriers, trains and shipping lines, with ripple economic impact across local to national economies. White collar jobs are being replaced by automation and AI in increasing numbers and range of roles, in including banking, investing, brokerage and insurance.
A Japanese insurance company, Fukoku Mutual Life Insurance in January replaced 34 claims workers (30% of the claims staff) with a $1.7m IBM Watson based AI investment (7). Other Japanese insurance companies and US companies have also started using AI to replace staff.
Where will the new jobs be? In engineering perhaps; banks are replacing their bankers in significant numbers with small numbers of engineers; in AR, VR, robotics, quantum computing? All these jobs are for the highly intelligent. Will there be a net creation rather than destruction of jobs? What are the new, relevant skills we will all need to succeed? What will robots, AI not be able to do? The potential impacts are more serious, calamitous though than loss of jobs, livelihoods, mass unemployment and global depression.
In 1996, IBM’s Deep Blue computer defeated world chess champion Garry Kasparov. In 2017, AlphaGo (without human defined rules), developed by Google DeepMind defeated Go world champion Lee Sedol. You can argue that great as these developments have been, they have been made using specific purpose AI. That argument fails, as general purpose, or modulizable AI is being developed. The intelligence levels and capabilities of AI are beginning to and will surpass humans. Humans have not had to compete against another species since the Neanderthals around 40,000 years ago. This unprecedented development in the history of humanity puts our future in peril. We cannot not fall behind AI. We will need evolve or we may perish. Will this evolution be to Elon Musk’s cyborg future?
Our destiny most definitely will not be in our hands, if as a 2016 Wall Street Journal article (4) noted, a sufficiently evolved AI views humans as having an intelligence level not dissimilar to that of a house plant.
There is a missing link here, as you will perceive. Until recently I had believed, from a philosophical rather than religious perspective, that AI could not make the leap from intelligence to understanding and self-awareness, consciousness, from the tool to the tooler. I no longer believe in the impossibility of AI self-awareness. I fear now a morality play that the Gods drop on us, to help us improve, learn lessons we are obviously not learning now.
While we don’t want a state of hysteria surrounding impacts of automation and AI, we cannot underrate the seriousness of our current situation.
The topic of impacts of automation should not be treated lightly, or with broad analogies in place of reasoned arguments supported by available evidence. Our futures and our children’s futures are at stake. We need to make decisions for our families and in our societies to ensure our future freedoms. We need to advocate for policy changes. We need to support policy makers to understand the portentous nature of these impacts, to discuss, debate and prepare for these changes.
CALL TO ACTION
You’ve read this far. Thank you. You now have a few options open –
- You can ignore this warning, or create an excuse — like too busy, maybe next month. This alleviates you of the need to take action. Your accountability and responsibility are not absolved.
- You can say the prognosis is faulty, or the evidence presented is not conclusive, there are gaps, holes, even chasms evident. Tell me where the prognosis is faulty, where more evidence is required.
- You can say you’re listening but not persuaded. This is good. Tell me so.
- Or, you can agree we need take action now, today, and say, “Where do I sign up, what’s next ?”
What do you say?
- From the title of this essay, it may appear that if in agreement, all is gloom ahead. That is not necessarily the case, if you will bear with me through my next dispatch. We have free will to impact, to create our future.
- Please treat this essay in the agile sense as an early, 2nd iteration. I haven’t yet fully supported my prognosis, or achieved my goal for this essay. I had to get it out tonight. Further support will come with my next dispatch. This is my first non-fiction published article in recent times. If you know of a good editor please tell me.
- I welcome your feedback, critiques, thoughts and in general developing of the discussion. If you can recommend any seminal articles, please link in the comments.
- Parts of this essay are taken from my comment on the Pulse article (5) published Feb 23, on LinkedIn.
- Jan Marsli provided valuable feedback to my LinkedIn comment. I’ll elaborate on the points Jan raised in my next dispatch.
- Thanks to Medium for providing this platform for publishing articles and supporting discussion and debate.
- An Open Letter: Research Priorities for Robust & Beneficial Artificial Intelligence. Future of Life Institute, Jan 15. https://futureoflife.org/ai-open-letter/
- A warning from Bill Gates, Elon Musk, and Stephen Hawking, Quincy Larson, Feb 19, 2017. https://medium.freecodecamp.com/bill-gates-and-elon-musk-just-warned-us-about-the-one-thing-politicians-are-too-scared-to-talk-8db9815fd398#
- Artificial intelligence: ‘We’re like children playing with a bomb.’ Tim Adams, Guardian, June 12, 16. https://www.theguardian.com/technology/2016/jun/12/nick-bostrom-artificial-intelligence-machine
- Machines that will think and feel. Artificial intelligence is still in its infancy — and that should scare us. David Gelernter, Wall Street Journal, Mar 18, 2016 (may be paywall). https://www.wsj.com/articles/when-machines-think-and-feel-1458311760?mod=e2tw
- Why Automation Will Create, Not Destroy, Jobs. Anurag Harsh, LinkedIn Pulse, Feb 23, 17. Also a Youtube video linked in article. https://www.linkedin.com/pulse/why-automation-create-destroy-jobs-anurag-harsh
- Artificial Intelligence will empower us, not exterminate us. Oren Etzioni, YouTube, TEDxSeattle, Feb 2, 17. https://youtu.be/H9Ddd1H9uLE
- Japanese company replaces office workers with artificial intelligence. Justin McCurry, Guardian, Jan 5, 17. https://www.theguardian.com/technology/2017/jan/05/japanese-company-replaces-office-workers-artificial-intelligence-ai-fukoku-mutual-life-insurance