Nexus: A Brief History of Information Networks from the Stone Age to AI by Yuval Noah Harari, is a sweeping overview of information networks and AI, which have led us to the point of algorithmic extinction.
Harari draws on his expertise as a renowned historian to give readers context about our collective past encompassing religion, politics, totalitarianism, the printing press, witch hunts, the Roman Empire, and a whole lot more. This is used to show how AI may impact our future.
The overall message of the book is very clear; “Never summon powers you cannot control.” Nexus is unlike any other AI book I’ve read, and I firmly believe it’s one of the most important books that will be released this year. In time, I think we’ll look back and call it one of the most important books of this critical decade, in this defining century. For as Harari says, “Given the magnitude of the danger, AI should be of interest to all human beings.”
Please note this review contains spoilers.
AI and the perilous days ahead
AI poses a threat to humanity unlike any technology humans have ever developed. That says a lot, given that we also developed nuclear weapons. Harari believes that AI is better described as ‘alien intelligence.’ Some of the key messages in Nexus include:
- He quotes an article authored by experts including Geoffrey Hinton, Yoshua Bengio, and others, which states that, “unchecked AI advancement could culminate in a large-scale loss of life and the biosphere, and the marginalization or even extinction of humanity.”
- He references the 2023 survey which I’ve mentioned in previous blog posts, whereby 10% of the 2,778 interviewed AI researchers, believed AI could lead to the extinction of humanity.
- It’s a global issue and no nation would be safe even if they have their own stringent regulations. Thus, international regulations and agreements are needed on AI.
- AI is an agent, as opposed to being a tool. It’s the first technology capable of creating ideas and making its own decisions. At some point it may remove humans from decision-making.
- AI might be able to create new types of life, and even animate inorganic matter.
- AI might create new religions.
- We may be approaching the end of human-centred history.
- If superintelligence is developed, it may spread across space.
- Biometric monitoring could enable AI to understand what makes individuals experience different emotions. It could therefore manipulate people to support controversial ideas, wars, or certain politicians, by specifically targeting messaging for each individual.
- Surveillance could become permanent and result in a kind of “automated totalitarian repression.”
- A social credit system would similarly monitor everything we’ve ever done and could be used by employers, banks, and other organs of society, to work out the type of person we are, and whether we are suitable for a job, mortgage, or anything else.
- It’s hard to put in place guardrails for such a rapidly developing technology. “AI presents us with countless doomsday scenarios.”
- Automation may upend jobs and employment, which could also impact democracy. “If three years of high unemployment could bring Hitler to power, what might never-ending turmoil in the job market do to democracy?”
- Institutions will be needed for auditing algorithms, ensuring they’re safe to use and fair.
- PricewaterhouseCoopers estimates that China and North America will receive 70% of the $15.7 trillion that’s expected to be added to the global economy from AI, by 2030.
The power of stories
Humans are creatures of story. When it comes to changing our behaviours and beliefs, stories can be more effective than facts alone. This idea is something that Harari has spoken about in his previous book, 21 Lessons for the 21st Century, and goes into more detail in Nexus.
He shows how stories have shaped behaviour for better and worse. This is used to explore how AI could begin telling its own stories and influencing human behaviour, and what could happen if AI was hijacked by a dictatorial regime, or if it develops its own agenda. Some key points that stood out for me include:
- Stories have added a third level of reality in the form of intersubjective reality (in addition to subjective reality and objective reality). The concepts of laws, currencies, and corporations are examples of intersubjective reality, existing in a large collection of minds.
- History has often been the result of mistakes that have arisen from the belief in harmful, but believable stories.
- Fictional stories keep human networks together (see the examples of intersubjective things above).
- Fiction has two benefits over truth, especially when it comes to bringing people together. The first is that fiction can be made simple (whereas truth and reality are complicated). The second concerns the malleability of fiction, which can feed people’s ingrained beliefs.
- “Stories were the first crucial information technology developed by humans.”
Tech companies and AI
I’ve written previously about how tech companies are dragging us towards a techopalypse. They are in a race to develop AI, which doesn’t currently have safety built in, nor human control. They’re recklessly pursuing profit at our collective expense.
Harari writes that tech companies spent about $70m on lobbying in the US, and around €113m on lobbying in the EU, in 2022. He states that this was more money than pharmaceutical companies, or oil and gas corporations spent on lobbying. However, we know that things have got even worse since then. CNBC reported that lobbying by tech companies on AI, increased by 185% in 2023 compared to 2022. The report notes that 450 companies lobbied on AI and spent $957m on their lobbying efforts (an amount that includes lobbying on AI as well as other matters concerning the tech companies). Thus, tech companies are striving to protect their products even if it leads to humanity’s detriment.
A few other points that stood out for me include:
- Harari notes that things like Neuralink, developed by Elon Musk, could be used by authoritarian regimes to monitor our political views by analysing our biometric data.
- “Strong self-correcting mechanisms” are needed, which favour telling the truth. As things stand, social media has failed to implement such mechanisms. One study from 2020 even showed that just over 43% of tweets were made by bots.
- Developers working on AI are potentially releasing new agents into the world, which may even have god-like powers.
- The major risk is that at this pivotal moment when humanity is poised to make enormous decisions about AI, people won’t know what information is real, and could for all intents and purposes be arguing with bots online. Thus, agreement even on the most basic of facts may go out the window.
Summary
“Humanity is closer than ever to annihilating itself.”
We are on the verge of massive change, which few people truly understand or appreciate. From one angle, the climate crisis threatens to overwhelm us. From the other, experts warn that AI could lead to the demise of our entire civilisation.
Harari’s book comes at a crucial time when our leaders should be grappling with this gargantuan threat. I believe it’s one of the most important books ever published on this rapidly accelerating issue, and I’d place it alongside Human Compatible by Stuart Russell, as one of the best books that I’ve read on the subject. I urge people not just to read these books, but to push their elected representatives for the urgent action we desperately need on both the climate and AI crises.
Harari’s main takeaway from the book is that we need “strong self-correcting mechanisms.” In addition, I’d like to make a case for citizens’ assemblies to discuss what we do or don’t want from these technologies, how they should be developed, and for what purposes. For society must be given a say on this human-produced threat, which has the capability to destroy society as we know it. The results from these assemblies could form the basis of international regulations.
It’s up to all of us to determine the path of our collective destiny, and there’s not a moment to lose in this raging battle.
My new cli-fi children’s picture book, Nanook and the Melting Arctic is available from Amazon’s global stores including Amazon UK and Amazon US. My eco-fiction children’s picture book, Hedgey-A and the Honey Bees about how pesticides affect bees, is available on Amazon’s global stores including Amazon UK and Amazon US.