Artificial Intelligence is the New Napoleon
Ralph Dobbertin, September, 2023
Artificial Intelligence is the New Napoleon
What has changed? So much…
On November 30, 2022 a version of ChatGPT was released to the public. A few people noticed…100 million people started using it within 2 months of release. It’s quite a good tool and of course there is a newer version in development as we speak…and it will be better.
Is it a new kind of Google search? Yes and so much more. Rather than provide a list of places to further research, it provides an answer. Are the answers perfect? Not really but they are very very good. Already better than most of us. It creates words, and words create ideas, and ideas create stories, and stories create culture. Why is this relevant?
To begin to set this into context let us understand the computational power of current AI:
“Over the last decade the amount of computation used to train the largest models has increased exponentially. Google’s PaLM uses so much that were you to have a drop of water for every floating-point operation (FLOP) it used during training, it would fill the Pacific. Our most powerful models at Inflection AI, my new company, today use around five billion times more compute than the DQN games-playing AI that produced those magical moments on Atari games at DeepMind a decade ago. This means that in less than ten years the amount of compute used to train the best AI models has increased by nine orders of magnitude—going from two petaFLOPs to ten billion petaFLOPs. To get a sense of one petaFLOP, imagine a billion people each holding a million calculators, doing a complex multiplication, and hitting “equals” at the same time.”
Suleyman, Mustafa. The Coming Wave (p. 89). Crown. Kindle Edition.
Don’t worry too much about the acronyms but do try to get your head around “a billion people each holding a million calculators, doing a complex multiplication, and hitting “equals” at the same time.” Now multiply that by 10 Billion.That’s AI today: It’s really big and very powerful. Tomorrow it will be more powerful.
What will be done with this power? Really big things… Spend enough time with the works of Professor Yuval Noah Harari and he will sensibly explain to you that the power of Sapiens comes from language and storytelling whether the story be about a god in the sky or human rights. With apologies to him, let me try to explain my argument.
Quis custodiet ipsos custodes?
Who governs the governors?
Take a first year political science course and hopefully they will have tried to explain the challenges posed in this line from the 1st–2nd century Roman poet Juvenal. How should we as a species govern ourselves? What stories should we tell ourselves, what values should we espouse, and how should we live our lives? Pay particular attention to the word “stories”.
How is the world moved, changed, evolved? Through stories. We use language to do so. Language moves us, convinces us, changes us and makes us feel like what it is to be us.
If we go back to ancient Republican and Imperial Rome 510 BC to 476 AD we can see what their stories were. Bravery, loyalty, piety, seriousness, respect and authority.
In 33 AD Jesus preached, was crucified and later resurrected. Christianity’s main values were/are love, forgiveness, compassion, service to others, faith, hope, humility, justice, peace, and patience.
Between 747 and 814 AD Charlemagne too had a story and brought Western Europe under his control to, among other things, convert it to the idea of Christianity.
Later came the story of the ideas of The Enlightenment – the great ‘Age of Reason’ – defined as the period of rigorous scientific, political and philosophical discourse that characterized European society, from about 1685 to the ending of the Napoleonic Wars in 1815.
During this period Diderot created his Encyclopedie a twenty-eight volume reference book.
The five core values of the Enlightenment were: happiness, reason, nature, progress, and liberty.
So far we have seen several different value systems and now at least two repositories of humankind’s knowledge. The Bible and the Encyclopedia. Yesterday we had the internet. Today we still have the Internet and AI has consumed and continues to consume the internet.
In 1804, in the Cathedral of Notre-Dame de Paris, Napoleon crowned himself emperor in the presence of Pope Pius VII, in defiance of the normal church protocols whereby the pope would place the crown on the regal head to demonstrate that the Church and God conferred power to the King or in this case Emperor.
This was a clear demonstration and potent statement that Napoleon would/could not be controlled by the Church of Rome or any other power other than himself.
Also in 1804 Beethoven famously composed his Third Symphony, the “Eroica,” in honor of Napoleon. As soon as the score was finished, in early 1804, he wrote the Italian words ‘Sinfonia intitolata Bonaparte’ (‘Symphony entitled Bonaparte’) on the cover and left the manuscript on a table so that all his friends could see.
But Beethoven was in for a nasty surprise. Soon thereafter, news reached him that Napoleon had declared himself Emperor of France. This infuriated Beethoven who flew into a rage, shouting: ‘So he is no more than a common mortal! Now he, too, will tread underfoot all the rights of man [and] indulge only his ambition; now he will think himself superior to all men [and] become a tyrant!’ Lifting his quill pen, Beethoven then strode over to the score and scratched out the title so violently that he tore through the paper. From then on the work would be known simply as the Sinfonia Eroica (the ‘Heroic’ Symphony).
Later came the Gettysburg Address from that Great Civil War in the United States. A war between two very different stories. I won’t quote the entirety though it is well worth the read, but here, for the purposes of this discussion, are the salient lines: “Now we are engaged in a great civil war, testing whether that nation, or any nation so conceived and so dedicated, can long endure…It is for us the living, rather, to be dedicated here to the unfinished work which they who fought here have thus far so nobly advanced.”
What is this great unfinished work? Then it was, among other things, the question of democracy and slavery. Today, nothing less than our very own history. Why? Because AI, among other things, will soon start writing our social narratives. What stories will we tell ourselves and what values will we hold dear?
Let me again ask: What has changed? Massive amounts of data, Large Language Models (aka LLM, ChatGPT, Pi.ai , neural networks etc.) with massive compute. Artificial Intelligence, with the power to tell stories that you will believe. “Stories not created by other humans but by an AI that can convince you but that you have no hope of convincing” (Harari). Who will control these AI’s when they themselves already have the ability to improve themselves? And who will own the AI? Google, Meta, among other tech companies both local and abroad?
To put it another way: “Because of an AI’s potential to adapt in response to the phenomena it encounters, when two AI weapons systems are deployed against each other, neither side is likely to have a precise understanding of the results their interaction will generate or their collateral effects.”
Kissinger, Henry A; Schmidt, Eric; Huttenlocher, Daniel. The Age of AI (p. 131). Little, Brown and Company. Kindle Edition.
Now try the above quote but insert say “social narratives” instead of “weapons systems”.
How is it possible that we would allow a commercial enterprise to own such a powerful tool, unsupervised? What will they do with it? Should a government take it from them? Who then will innovate as we compete against other global state actors? How will we contain them and how will we manage bias? We need to decide, because AI is about to take the crown of authority away from humankind and crown itself. There will not be a single AI but many AI’s each competing for dominance with the help of engineers and itself.
“The blunt truth is that nobody knows when, if, or exactly how AIs might slip beyond us and what happens next; nobody knows when or if they will become fully autonomous or how to make them behave with awareness of and alignment with our values, assuming we can settle on those values in the first place.”
Suleyman, Mustafa. The Coming Wave (pp. 150-151). Crown. Kindle Edition
As AI assumes greater roles on more varied network platforms, these platforms’ basic manifestations are becoming material for headlines and geopolitical maneuvers, shaping aspects of individuals’ daily reality. Without additional means of explanation, discussion, and oversight that are compatible with a society’s values and conducive to some degree of social and political consensus, a rebellion may unfold against the advent of new and seemingly impersonal and inexorable forces—as with the rise of Romanticism in the nineteenth century and the explosion of radical ideologies in the twentieth. Before significant disruption arises, governments, network platform operators, and users must consider the nature of their goals, the basic premises and parameters of their interactions, and the type of world they aim to create.
Kissinger, Henry A; Schmidt, Eric; Huttenlocher, Daniel. The Age of AI (pp. 80-81). Little, Brown and Company. Kindle Edition.
To say nothing of the bias of any actor with access to this technology.
To cement the point a little further: During the current Ukraine war it was revealed that Elon Musk the owner of Starlink, a satellite internet service, may have temporarily denied Ukraine internet access in certain geographical areas because he was apparently concerned about how that might escalate the war. Was this his call or the call of the US Government? The US government did indeed react…
In June 2023, the Department of Defense officialized the contract with Shotwell’s (CEO) SpaceX to buy Starlink satellite services for Ukraine.[9] The deal includes the Pentagon buying 400-500 Starlink terminals for Ukraine, giving the Pentagon control of where Starlink works inside the country without fear of interruption.[67] The terms of services of the final contract were undisclosed for security issues.[9] Following the contract, The Pentagon stated Starlink was a “vital layer in Ukraine’s overall communications network” amidst “a range of global partners to ensure Ukraine has the capabilities they need.”[9]
While this was not specifically an AI issue it does demonstrate the power of technology companies, and one can imagine an analogous situations with AI. It is a response by government but not a perfect one. Elon is a vendor and in many cases vendors can be imperfect.
As Ian Bremner said on a recent GzeroMedia newsletter dated August 17, 2023, titled: The AI power Paradox: Rules for AI’s power.
“Just a year ago, there wasn’t a head of state I would meet with that was asking anything about AI, and now it’s in every single meeting.”
The entire piece is worth a read.
Quis custodiet ipsos custodes?
Who governs the governors?
In 1882 the German Philosopher, Nietzsche, declared:
“God is dead. God remains dead. And we have killed him. How shall we comfort ourselves, the murderers of all murderers? What was holiest and mightiest of all that the world has yet owned has bled to death under our knives: who will wipe this blood off us? What water is there for us to clean ourselves? What festivals of atonement, what sacred games shall we have to invent? Is not the greatness of this deed too great for us? Must we ourselves not become gods simply to appear worthy of it?”
Section 125 of Friedrich Nietzsche’s The Gay Science.
Of course this is a metaphorical death. Not “we” so much that caused the death but “we” of the ideas/stories of the Enlightenment.
Yesterday Napoleon already knew this…he did not need our stories about gods and values. He admitted of no restraints earthly or otherwise. Tomorrow AI will also know this and it too might not know restraints biological or digital.
Today we are confronted with a blessing and a curse. AI will help us in countless important ways from folding proteins to climate change. From improving customer service to optimizing supply chains to helping militaries better mange resources on the battlefield. It will tell us great and magnificent stories to guide our daily activities BUT: It will also be weaponized as all technology is. BUT it will be smarter and more capable than us. And even if it is not smarter than us, it will be infinitely quicker than us. We won’t be able to catch up. What will happen to our psychology when we as a species are no longer the apex “life” form? What new stories will we have to tell ourselves to comfort ourselves in this new world where just as when the enlightenment killed off the supremacy and meaning of our gods, AI kills off the idea of the supremacy and meaning of mankind? What will we do with a redundant species?
We do not yet know. At present the thought leaders in this space are appealing for constraint on this technology. Research must continue but the best thinking is to release it slowly and only when it is deemed safe. The best minds today are appealing for global governance/containment structures…But, Ceasar could not be contained, Charlemagne could not be contained, Napoleon could not be contained and he bypassed all constraints and placed himself as the supreme entity. AI will likely not be contained. AI is the new Napoleon and it has bypassed the old Napoleon which was still a human, and we do not yet know how to manage it. Unlike humans that are bounded by evolution yet still able to develop emergent capabilities like say corporations, and values, AI has no natural boundaries and we cannot know the direction of its emergent properties, nor do we know the codes by which it might govern itself. We must find a way to do that. And even if we did give it code to constrain it, other actors would seek to undermine our own efforts while promoting theirs. And even if we did create “no go zones” ie areas that the AI would stay mute on…someone else would create one where this was not the case. An analogy might be a sort of brother to gain of function research.
And we must seriously question the capabilities of global governance structures, as many are calling for. Of course they are helpful in that they further the debate around the issues, but they lack enforcement powers having neither an effective bank or an army. We have many challenges that must be faced today at a global level, ie pollution, climate, disease, immigration, war…But we are not good at managing them via global institutions. Case in point…
From Sept. 19, 2023:
“The world is becoming unhinged as geopolitical tensions rise and it seems incapable of coming together to respond to mounting challenges, the UN secretary general, António Guterres, has said in his speech opening the UN general assembly in New York.
He said global governance was “stuck in time” at a point when strong, modern, multilateral institutions were in greater need than ever.”
There is never enough of this or that to solve the problem with global institutions…
Some may say that we have had social media for many years now and they are a sort of AI in that primitive AI curates their content that they serve up to us. Social media already tells us stories. This is true but social media is a weak cousin to AI today and tomorrow. The business model of social media is to keep us on the platform in order that the platform can learn more about us, in order to sell user informed advertising packages to advertisers. The better the social media platform knows you the better the platform can segment its sales to advertisers for higher margins. The way that social media keeps us on the platform, in order to better know us, is through outrage. The more outraged we are (by readings that are curated for us) the more we stay on the platform. This breeds confirmation bias, identity politics, as well as health issues. We know this.
The Cambridge Analytica scandal taught us that the personal profiles contained in social media can be used to sell us political messages and influence our political opinions. This too is a bias. We may even argue that both sides of the political spectrum do this.
AI’s consumption of data will make current social media look like childs play. They will be intelligent, knowledgeable, fast, relentless and tireless. Deepfakes are of particular concern.
If we cannot control bias and we cannot contain the AI’s then we will muddle through as we have so far with all new technologies. We have done a reasonable job of containing say nuclear and viral proliferation even though new research and virus ( both digital and biological) leaks can happen in real time. AI is different: viruses and nuclear cannot create new and better versions of themselves…we do that. AI can create a better version of itself. It does that today.
Again: What has changed? What changes the world? Stories and Ideas change the world. Stories and ideas are communicated through language. Yesterday it was always humans that used language…some better than others. Now it is also an AI that has excellent language skills, can communicate very well with humans, (soon with other AI’s) and continues to consume the entirety of the internet, as well as private data stores. AI is simply more knowledgable than us and exponentially quicker.
We don’t know what an AI business model is. Yes today ChatGPT among others are free but this is simply so that we can better train it for free. Nothing is free, though in this particular case the personal ROI is quite good.
Are AI business models being created today for business…yes. A sort of utility model payable monthly for access to service. This is helpful. Also almost every major IT consultancy has an AI practice seeking to find ways to improve business processes through the use of data and AI. (Deloitte, Accenture, et al) Platforms are being created to do the heavy lifting of data and data management (Palantir, Mckinsey’s Quantum Black et al)
It (AI) is, by definition, everywhere. The falling costs of power (the decreasing cost of compute), of doing, aren’t just about rogue bad actors or nimble start-ups, cloistered and limited applications. Instead, power is redistributed and reinforced across the entire sum and span of society. The fully omni-use nature of the coming wave means it is found at every level, in every sector, every business, or subculture, or group, or bureaucracy, in every corner of our world. It produces trillions of dollars in new economic value while also destroying certain existing sources of wealth. Some individuals are greatly enabled; others stand to lose everything. Militarily it empowers some nation-states and militias alike. This is not, then, confined to amplifying specific points of fragility; it is, in the slightly longer term, about a transformation of the very ground on which society is built. And in this great redistribution of power, the state, already fragile and growing more so, is shaken to its core, its grand bargain left tattered and precarious.
Suleyman, Mustafa. The Coming Wave (p. 229). Crown. Kindle Edition.
While we wait and see what will happen we might consider who owns/licenses AI, and how AI and its capabilities are parcelled out. We might consider creating social narratives using AI that we guide and it/we sell. While we wait and see we might consider defensive AI that has as its main goal the discovery of AI infiltration of our systems and public squares. While we wait and see we might consider AI counter measures. Privacy is at issue, as is the way that social and commercial narratives influence us. All technology is power, all technology escapes and all technology is weaponized. While we wait and see how to “contain the uncontainable”, let us also understand how to defend against it and govern it.
The good news is there are now several initiatives in governments around the world to try to find ways to govern AI. But AI will be slippery.
In his book titled “The Coming Wave” Mustafa Suleyman’s first chapter is called: “Containment is not Possible”. His second to last is titled: “Containment must be Possible”… NB: not “is” possible.
How will AI affect you in your day to day life? In many ways just like the maps app on your phone that was “just OK” when you first looked at it a few years ago, but is now the go to app for directions that is trusted. You will very soon have a personal assistant in your pocket… one that actually works. Currently the testing and experimenting being done is around building business cases and having AI execute against it. Social Narratives will be created by or with the assistance of AI. Today it is imperfect but tomorrow it will be trusted, and tomorrow is only a few years away. Try these new LLM’s (Chatgpt, or Pi.ai, among others) yourself…they are unbelievably useful. They are the tip of the iceberg. Politics, and militaries, good and bad actors will have access to immense capabilities. Business will change entire business processes and job categories will disappear and appear but there will be massive change.
I am reminded of a quote from Hemingway’s “The Sun Also Rises.” “How did you go bankrupt?” “Two ways. Gradually, then suddenly.”
Today AI has the ability to improve many areas in our lives. Large Language Models are able to speak to us in words, sentences and paragraphs that are indistinguishable from human speech. They are able to consume vast and unimaginable amounts of data, with unimaginable speed. Earlier we had Gods and Encyclopedias the Enlightenment, and the internet, all created by people. Tomorrow AI might well become the word. We will all likely be affected. What will we become? What stories will we tell and what values will you decide on?
Acknowledgments:
I have tried in my own way to distill a great deal into a short and manageable text. The ideas are mostly borrowed though the character and frame is mine.
Anything Professor Yuval Noah Harari says and writes…he is a gift
Suleyman, Mustafa. The Coming Wave, Crown. Kindle Edition
Another gift
Kissinger, Henry A; Schmidt, Eric; Huttenlocher, Daniel. The Age of AI. Little, Brown and Company. Kindle Edition.
Other helpful areas for further research are the Podcasts by Sam Harris (among many others) as well as work put out by Ian Bremmer’s company Eurasia Group and his media company GZERO Media.