The Age of AI and Our Human Future by Henry A. Kissinger, Eric Schmidt, Daniel Huttenlocher

The authors say: "No expert, no matter his or her field, can single-handedly comprehend a future in which machines learn and employ logic beyond the present scope of human reason".

"Humans are creating and proliferating nonhuman forms of logic with reach and acuity that... can exceed our own and humans could not articulate precisely why it worked".

We are building machines more 'intelligent' than us, that do things we can't explain. They are also exposing us to new forms of intelligence and 'relationship' pathways that we never thought of.

AI is already making major decisions about our life and future(e.g. in research, recruitment and hiring, insurance claims, social media).

A game of chess

"quiet revolution occurred. AlphaZero, an artificial intelligence (AI) program developed by Google DeepMind, defeated Stockfish — until then, the most powerful chess program in the world."

"Prior programs had relied on moves conceived of, executed, and uploaded by humans — in other words, prior programs had relied on human experience, knowledge, and strategy... chief advantage against human opponents was not originality but superior processing power, enabling them to evaluate far more options in a given period of time."

"AlphaZero had no preprogrammed moves, combinations, or strategies derived from human play. AlphaZero’s style was entirely the product of AI training: creators supplied it with the rules of chess, instructing it to develop a strategy to maximize its proportion of wins to losses. After training for just four hours by playing against itself, AlphaZero emerged as the world’s most effective chess program."

"tactics AlphaZero deployed were unorthodox — indeed, original. It sacrificed pieces human players considered vital, including its queen. It executed moves humans had not instructed it to consider and, in many cases, humans had not considered at all. It adopted such surprising tactics because, following its self-play of many games, it predicted they would maximize its probability of winning."

AI developed "patterns of moves across vast sets of possibilities human minds cannot fully digest or employ."

"After observing and analyzing its play, Garry Kasparov, grand master and world champion, declared: “chess has been shaken to its roots by AlphaZero.”"

"the world’s greatest players did what they could: watched and learned."

AI the scientist

Authors give another example of AI being used by researchers at MIT to explore new antibiotics.

AI processed traits of 61,000 molecules, and identified one as a candidate antibiotic. Evaluating a molecule as potential antibiotic is a long and complex process. Conventional approaches are time-consuming and expensive.

AI did it faster and cheaper. But the surprising thing was:

"AI identified relationships that had escaped human detection — or possibly even defied human description... it detected new molecular qualities — relationships between aspects of their structure and their antibiotic capacity that humans had neither perceived nor defined. "

Defying explanation

"AI did not just process data more quickly than humanly possible; it also detected aspects of reality humans have not detected, or perhaps cannot detect."

Since human researchers cannot perceive or define new aspects of how it worked, "humans could not articulate precisely why it worked".

"each program had mastered its subject differently from humans. In some cases, it obtained results that were beyond the capacity of human minds — at least minds operating in practical time frames — to calculate."

"humans remain uncertain to this day how the programs achieved their goals"

We can't explain what we can't understand and don't even have the vocabulary for.

"The advent of AI obliges us to confront whether there is a form of logic that humans have not achieved or cannot achieve, exploring aspects of reality we have never known and may never directly know."

Undoing Enlightenment

The age of Enlightenment in Europe brought about a change in people's mindset. We could rationally explain what was going on in the world around us.

All natural, physical, even emotional phenomena could be rationally explained in the age of reason.

There were no black boxes.

Humans feel in control when we can explain what is going on around us. This is turn leads to less stress as we can predict the next step, and don't get anxious.

A similar analogy is delays on flights, trains, roads. The more we know and understand, the less stressed we are.

AI will undo the age of Enlightenment. Our world(and maybe even our life) will be controlled by AI. It will take all the decisions. We might not even have the capability to understand them. AI may even chose not to tell anyone anything. E.g. why it chose to kill us in a self driving car!

The result will be humans merely becoming even more insignificant pawns in a world run by AI, for the benefit of AI's corporate and government masters.

"During the Enlightenment, René Descartes’s maxim, Cogito ergo sum (I think, therefore I am), enshrined the reasoning mind as humanity’s defining ability and claim to historical centrality."

"Now the partial end of the postulated superiority of human reason, together with the proliferation of machines that can match or surpass human intelligence, promises transformations potentially more profound than even those of the Enlightenment. "

In the service of mankind

AI is a technology. It can be put to good use.

The book has example of how AI was used to "optimize the cooling of Google’s temperature-sensitive data centers. Although some of the world’s best engineers had already tackled the problem, DeepMind’s AI program further optimized cooling, reducing energy expenditures by an additional 40 percent—a massive improvement over human performance".

AI as the disrupter

"failing to apply that AI, at least as an adjunct to human efforts, may appear increasingly as perverse or even negligent."

If your business or domain does not apply AI, you are ripe for disruption, and open to accusation of negligence and incompetence.

This probably explains why there is explosion of investment in AI in every sector. Most don't understand it, but they do their best in order not to be accused of negligence and incompetence.

Everyone is on the bandwagon, few know how to play the band.

It is imposter syndrome on steroids.

The Process: the God of the corporate and government drones

Corporations and government already don't have any transparency. They want to know everything about us, while tell us nothing about their activities. It is a one-way mirror.

They hide their decisions under the magic word of 'Process'. Their decisions are the outcome of 'The Process' they followed. You might find minimal information about this 'Process' that made decisions about our life.

With AI, even the researchers will probably not have the vocabulary to explain why AI did what it did. 'The Process' will now be even more opaque.

Corporations and government will be even more opaque, unhelpful, arbitrary, punitive, and intrusive. This will probably be the change that will affect us most in our day-to-day life.

It will be tyranny of AI. Do as you are told, and random punishments for 'bad behavior'.

Shoot first, explain never

AI in self driving car can decide to kill its driver to prevent a bigger accident. This is well known use case of self driving car and extensively debated.

Other bigger use case is the use of AI in the battlefield. "what if AI recommended that a commander in chief sacrifice a significant number of citizens or their interests in order to save, according to the AI’s calculation and valuation, an even greater number?"

This has already been observed in game of chess where AI chess program decided to sacrifice the queen early on.

Will humans be allowed to override AI's decisions?

Will humans be allowed to override AI's decisions 80% of the way through? Will they even want to?

How will humans pick up the pieces after overriding AI? What will humans suggest next when they don't understand how we got here?

What if AI recommends killing the military General as it sees bad leaders as threat to victory?

Imagine being asked to play a chess game mid-way after the queen and other key pieces have been scarified.

"If we are unable to fathom the logic of each individual decision, should we implement its recommendations on faith alone? "

Should we trust AI on faith? Should we have faith that corporations can do no wrong and can do no evil.

The age of enlightenment taught us to believe in reason and logic. Now we will be asked to have faith again, not in God or ourselves or other humans, but a machine made by a corporation.

AI for military use

"A race for strategic AI advantage is already taking place, particularly between the United States and China and to some extent Russia".

"The paradox of an international system is that every power is driven to act — indeed must act — to maximize its own security. Yet to avoid a constant series of crises, each must accept some sense of responsibility for the maintenance of general peace. And this process involves a recognition of limits. The military planner or security official will think (not incorrectly) in terms of worst-case scenarios and prioritize the acquisition of capabilities to meet them. The statesman (who may be one and the same) is obliged to consider how these capabilities will be used and what the world will look like afterward."

Our future depends on having wise statesmen with "recognition of limits" while planning for "worst-case scenarios and prioritize the acquisition of capabilities to meet them".

Kissinger writes in his book 'Leadership':
"no country has persuaded itself actually to use them – even in conflict with non-nuclear countries. As previously described, both the Soviet Union and the United States accepted defeat at the hands of non-nuclear countries without resorting to their own deadliest weapons. "

"impact of revolutionary technology makes the full application of these weapons cataclysmic while rendering their limited use difficult to the point of unmanageability... No diplomacy has yet been invented for threatening their use explicitly without the risk of preemption in reply."

Nuclear, Cyber, AI Weapons are necessary but nations that possess them have to constrain themselves because of Mutually Assured Destruction(MAD), political suicide, and international isolation.

A bigger challenge is automating AI specially in the domain of defense.

The authors admit that "Ambiguity inherent in the domain — combined with the dynamic, emergent qualities of AI and the ease of dissemination — will complicate assessments."

"In the AI age, long-held strategic logic should be adapted. We will need to overcome, or at least moderate, the drive toward automaticity before catastrophe ensues. We must prevent AIs operating faster than human decision makers from undertaking irretrievable actions with strategic consequences."

The question is unanswered: if we automate military weapons using AI, AI is likely to suggest actions based on changing military landscape that military planners and strategists cannot explain. Do we override AI, defeating the purpose of automation. Or do we override AI leaving us in a position we can't explain. Where do we go when we override AI's decision to attack a certain target.

This is not a theoretical scenario. As NATO enlarges and comes up right to Russia's borders, automation will be enabled to ensure fast response in case of sudden strike. Same scenario will exist in South China Sea.

As the authors say, automating AI "will complicate assessments".

Meanwhile, "The military planner or security official will think (not incorrectly) in terms of worst-case scenarios and prioritize the acquisition of capabilities to meet them."

The growth in AI for military, and automation of war using AI will continue.

An AI arms control treaty is required, but unlikely since nations are dismantling even the older already negotiated nuclear and conventional treaties.

Restricting AI

"The cautious may seek to restrict AI, confining its use to discrete functions and circumscribing when, where, and how it is used. Societies or individuals may reserve the role of principal and judge for themselves, relegating AI to the position of support staff. However, competitive dynamics will challenge limitations, "

"If AI enables a bureaucrat, architect, or investor to predict outcomes or conclusions with ease, on what basis would he or she not use it? Given the pressures for deployment, limitations on AI uses that are, on their face, desirable will need to be formulated at a society-wide or international level."

Restricting AI might only work if it is restricted at international level, and the laws strictly enforced. This is unlikely.

Missing from the book

The book offers no suggestions or recommendations for the common man.

The authors recommendations are for government to setup a government body to oversee AI. This body will just eat up resources and do nothing.

The authors recommendations are for leaders to be aware of limitations of use of powerful weapons like AI, Cyber, Nuclear. These weapons cannot be used even in wars with less powerful nations. They have a role as deterrence, and take lot of resources to develop.

Living with AI

The impact of AI will be as big if not bigger than social media and technology in general as we created a new form of intelligence that we cannot explain, but it will govern us.

AI will save corporations and government money, and fulfill their narcissistic desire of monitoring and control. There will be competitive pressure. No company can ignore it for fear of accusation of incompetence or negligence. It will be impossible to escape AI in our current system.

AI is already deployed behind the scenes as well as in your face e.g. in cars, corporate systems, recruitment, finance, insurance, governments, digital currency, facial recognition in every CCTV, and smartphones.
The common man will see some good, lots of benign uses, and some truly awful uses of AI. The recommendations is to be aware, and avoid its use as much as possible. That can be said for any technology.

"The age of AI has yet to define its organizing principles, its moral concepts, or its sense of aspirations and limitations."

It likely never will. The wild-west free-for-all will continue. We are still waiting for organizing principles, moral concepts, sense of aspirations and limitations, and restrictions for social media and the data industrial complex after more than a decade of it.

The major players in AI space will be the same as in social media - large corporations with resources and former government bureaucrats as board members, and the Deep State.

Future of AI:
Some good: new antibiotics, saving energy in data centers.

Lots of useful and benign developments: better keyboards on smartphones, better noise canceling earphones.

Some truly awful: AI will be the new god of 'The Process' in corporations and government, making corporations and government even more opaque, unhelpful, arbitrary, punitive, and intrusive than they already are. It will be tyranny of AI. Do as you are told, and random punishments for 'bad behavior'.



Tweet

Related articles

System Fail

How an Economy Grows and Why It Crashes

External Links