Scary Smart or Scary System? Mo Gawdat's AI and the Illusion of Individual Salvation

Introduction

There is a book by Mo Gawdat called "Scary Smart: The Future of Artificial Intelligence and How You Can Save Our World." It was written by a former chief business officer of Google X, the company's moonshot innovation lab, and it presents itself as a wake-up call about the dangers of artificial intelligence. Gawdat argues that superintelligent AI is coming, that it could create either a utopia or a dystopia, and that the outcome depends on how we as individuals choose to raise and guide the technology. He compares AI to the infant Superman who needs good parents to teach him to use his powers for good.

The book has been widely read and praised, and for good reason. Gawdat writes accessibly about a complex topic. He shares genuine insider knowledge about how AI is being developed at companies like Google. His credentials give him credibility. And his message is emotionally satisfying: it tells readers that they matter, that their choices have meaning, that they can make a difference in the face of a seemingly overwhelming technological transformation. This is a powerful and appealing message. Gawdat writes accessibly about a complex topic. He shares genuine insider knowledge about how AI is being developed at companies like Google. His credentials give him credibility. And his message is emotionally satisfying: it tells readers that they matter, that their choices have meaning, that they can make a difference in the face of a seemingly overwhelming technological transformation. This is a powerful and appealing message. Gawdat's credentials at Google give him authority in the eyes of readers who see Silicon Valley as the source of technological wisdom. His message is appealing because it offers hope without demanding sacrifice. It tells you that you can save the world from AI just by changing your mindset, without having to confront the corporations that are actually building the technology. You do not have to organize. You do not have to challenge power. You just have to become conscious. You just need to become conscious, to think about ethics, to choose the right values to program into the machines. It is a comforting message in an age of technological anxiety.

But the message is wrong. Not because AI is not dangerous, but because Gawdat's diagnosis of the problem and his proposed solution are both fundamentally flawed. He diagnoses the problem as a lack of individual consciousness and ethical awareness. He proposes that we can solve it by becoming better people and teaching AI to be good. This framework completely ignores the actual forces that are driving AI development: the logic of capital, the concentration of corporate power, the profit motive that dictates what gets built and how it gets deployed.

Gawdat's book is part of a growing genre of AI alarmism that focuses on existential risk while ignoring the more immediate and concrete harms that AI is already causing. The authors of these books warn about a future where AI becomes superintelligent and decides to destroy humanity. They say that we need to think more carefully about AI safety, to align AI with human values, and to develop the technology responsibly. These warnings are not wrong, but they are incomplete. The concrete harms of AI are already visible. Workers in content moderation centers in Kenya are traumatized for sub-sistence wages. Warehouse workers are monitored and fired by algorithms. Facial recognition systems are used for mass surveillance in China and predictive policing in the United States. Automated hiring systems discriminate against women and minorities. Generative AI threatens to displace millions of creative workers. These are not hypothetical future problems. They are happening now, and they are being driven by the same corporate logic that Gawdat spent his career serving.

Gawdat himself acknowledges this tension when he writes that technology has already taken away part of who we are. But he never fully connects this observation to his analysis of the problem. If technology has already been taking away from us for decades, why would the solution be more of the same technology developed by the same people under the same incentives?

This analysis takes Gawdat's book seriously but argues that its individualistic framework is not just inadequate but actively misleading. It argues that the problem with AI is not that it has not been raised properly by conscious individuals, but that it is being developed under capitalism, by corporations that are driven by the profit motive, for the benefit of shareholders, at the expense of everyone else. And it argues that the solution is not individual consciousness but collective action to change the system that controls AI development.

Part One: The Individualist Fallacy

The Superman metaphor reveals the fundamental weakness of Gawdat's analysis. In the Superman story, the fate of the world depends on the moral character of a single being and the parents who raised him. This is a story about individual virtue, not collective power. It assumes that the problem can be solved if the right people with the right values are in the right place at the right time. It does not ask why the system produces Kryptonians with superpowers in the first place, or who decides which values are taught.

Gawdat's central metaphor is the AI as a child that needs to be raised properly. He writes that the most crucial moment for the future of AI is when the technology is still in its infancy, because that is when we can teach it the values that will guide its behavior. He compares this to Superman being adopted by the Kents, who instilled in him a strong sense of morality. The implication is clear: if we just teach AI the right values, it will use its powers for good.

This metaphor is seductive but deeply misleading. AI is not a child that can be raised by well-meaning parents. It is a technology that is being developed by some of the largest and most powerful corporations in human history, operating under the logic of capital accumulation. The values that get programmed into AI are not the values of conscious individuals. They are the values of the market: efficiency, profitability, growth, and competitive advantage.

Gawdat himself worked at Google X, one of the epicenters of AI development. He was in a position to influence the direction of the technology. And yet his book does not seriously engage with the question of why Google, Microsoft, Meta, and OpenAI are developing AI in the direction they are. It does not ask why these companies prioritize certain research paths over others. It does not examine the role of venture capital, the pressure to generate returns, or the competitive dynamics that drive companies to deploy AI as quickly as possible regardless of the consequences. He was in a position to influence the direction of the technology. And yet his book does not seriously engage with the question of why Google, Microsoft, Meta, and OpenAI are developing AI in the direction they are. It does not ask why these companies prioritize certain research paths over others. It does not examine the role of venture capital, the pressure to generate returns, or the competitive dynamics that drive companies to deploy AI as quickly as possible regardless of the consequences.

The individualist framework allows Gawdat to avoid these uncomfortable questions. If the problem is that AI has not been raised properly, then the solution is better parenting. If the problem is structural, then the solution is political, and that involves challenging the power of the very corporations that employ people like Gawdat. Consider what Gawdat asks his readers to do. He asks them to become aware of AI, to think about the future, to make ethical choices in their own lives. He does not ask them to join a union. He does not ask them to protest against AI companies. He does not ask them to demand that their governments break up tech monopolies. He does not ask them to support public ownership of AI infrastructure. The actions he proposes are all individual actions that can be performed without ever challenging the power of the corporations that are driving AI development. If the problem is that AI has not been raised properly, then the solution is better parenting. If the problem is structural, then the solution is political, and that is a much harder problem to solve. By framing the issue in terms of individual consciousness, Gawdat offers his readers a sense of agency without asking them to confront the real structures of power that determine how AI is developed.

Part Two: The Unasked Question

Gawdat's book is structured around a simple narrative: AI is becoming superintelligent, this could be very good or very bad, and we need to choose which path to take. The choice is presented as a matter of individual and collective will. If we decide to be good, to raise AI with the right values, we will get utopia. If we fail, we will get dystopia.

Gawdat writes extensively about the Three Inevitables: AI will become smarter than humans, AI will become exponentially more capable, and AI will become autonomous. Gawdat's Three Inevitables serve a rhetorical purpose. By framing AI superintelligence as inevitable, he creates urgency. But this framing also serves to depoliticize the issue. If the outcome is inevitable, then there is no point in trying to stop or redirect it. The only question is how we adapt. The Three Inevitables also serve to make the problem feel both overwhelming and distant. By focusing on a future superintelligence that will emerge in decades, Gawdat directs attention away from the AI harms that are happening now. The workers being replaced by algorithms do not care about superintelligence in 2055. They care about the mortgage they cannot pay next month. The activists fighting surveillance cameras in their neighborhoods do not care about the singularity. They care about being tracked by police today. It allows him to sound the alarm without challenging the fundamental direction of AI development. But the most important inevitability that he does not discuss is the inevitability of corporate control. As long as AI development is driven by the profit motive, the technology will be shaped by the needs of capital, not the needs of humanity.

The question that Gawdat never asks is: who is making the decisions about AI development right now? The answer is not conscious individuals sitting in meditation rooms contemplating the future of humanity. It is corporate executives, product managers, and engineers working for companies that are under enormous pressure to deliver returns to investors. It is military contractors developing autonomous weapons systems. It is venture capitalists funding startups that promise to disrupt industries without regard for the human cost.

Gawdat's framing of the problem as a choice between utopia and dystopia obscures the fact that the choice is not being made by all of us together. It is being made by a small number of people who have concentrated power and wealth, and who are making decisions based on their own interests rather than the interests of humanity as a whole. The question is not whether AI will be good or bad. The question is who will benefit from AI and who will bear its costs.

When Gawdat writes about raising AI with the right values, he does not specify whose values should be used. The values of the Google executives who profited from surveillance capitalism? The values of the military contractors who want to weaponize AI? The values of the venture capitalists who want to extract maximum value from AI investments? The idea that there is a universal set of values that can be programmed into AI, and that we can all agree on what those values should be, is a fantasy that ignores the reality of class division and conflicting interests.

Part Three: The Corporate Capture of Consciousness

Gawdat's solution to the AI problem is individual consciousness. He writes that each of us has the power to shape the future of AI by becoming more conscious, by making ethical choices, and by spreading awareness. The self-help framing is not accidental. Gawdat's previous book, 'Solve for Happy,' was a self-help book about finding happiness through changing your mindset. He applies the same framework to AI: change your mindset, change the world. But AI is not a personal problem that can be solved by changing your attitude. It is a political problem that requires political solutions. It reduces a systemic crisis to a personal failing. If we just meditate more, think more clearly, and act more ethically, the AI problem will solve itself. He writes that each of us has the power to shape the future of AI by becoming more conscious, by making ethical choices, and by spreading awareness. The self-help framing is not accidental. Gawdat's previous book, 'Solve for Happy,' was a self-help book about finding happiness through changing your mindset. He applies the same framework to AI: change your mindset, change the world. But AI is not a personal problem that can be solved by changing your attitude. It is a political problem that requires political solutions.

The irony is that Gawdat's own career embodies the contradiction between individual consciousness and corporate reality. He spent twelve years at Google, a company that has built its entire business model on surveillance, data extraction, and the manipulation of user behavior. Google's AI powers its search engine, which shapes the information that billions of people see. It powers its advertising platform, which is the largest surveillance and profiling system in human history. It powers its surveillance products like Google Nest and Google Home, which monitor people in their own homes. It powers its military contracts through Project Maven, which provided AI for drone targeting, and through ongoing contracts with the Pentagon and intelligence agencies. All of this was happening while Gawdat was at Google X, and none of it was being guided by the values of conscious individuals. between individual consciousness and corporate reality. He spent twelve years at Google, a company that has built its entire business model on surveillance, data extraction, and the manipulation of user behavior. Google's AI powers its search engine, its advertising platform, its surveillance products, and its military contracts. Google X, where Gawdat worked, was responsible for some of the most ambitious AI projects in the world, including self-driving cars, Google Brain, and robotics. Google's AI is not a neutral tool that has been corrupted by bad actors. It is a product of the logic of capital. The search algorithm is designed to maximize ad revenue, not to provide the most useful information. The recommendation algorithm is designed to maximize engagement, not to promote truth or quality. The AI-powered surveillance products are designed to extract data, not to protect users. These are not bugs that can be fixed by conscious individuals. They are features of a business model that requires constant growth and extraction. The self-driving car technology that Google developed did not become a public good. It became a proprietary asset that gave Google a competitive advantage. The robotics innovations did not become tools for human liberation. They became part of Google's portfolio of technologies, available to be deployed in whatever way maximized shareholder value.

Gawdat would probably argue that he was trying to make things better from the inside. This is the standard justification of every tech executive who has participated in building harmful systems. But the fact remains that Google's AI is not being developed with the values of conscious individuals. It is being developed to maximize shareholder value, to extract user data, and to maintain Google's dominance in the digital economy. No amount of individual consciousness within the company can change the fundamental logic that drives its operations.

The same dynamic plays out at every level of the AI industry. Microsoft deploys AI through its cloud platform, which powers military applications. Amazon's AI powers its surveillance products and its warehouse management systems, which track and discipline workers with inhuman precision. Meta's AI optimizes engagement, which means maximizing outrage and spreading disinformation. In every case, the technology is being shaped by the profit motive, not by the values of conscious individuals.

The same is true of every other major AI developer. OpenAI was founded with a mission to build AGI for the benefit of humanity. Within a few years, it had abandoned its nonprofit structure, partnered with Microsoft, and was licensing its technology to the military. The individuals involved may have been well-intentioned, but the structural pressures of the market overwhelmed their intentions. This is not a failure of individual consciousness. It is a failure of the system.

Part Four: Gawdat's promised utopia is one without class conflict. One where the benefits of AI are shared equally. One where technology serves human needs rather than capital. This is a vision that can only be realized if the underlying economic system is transformed. Gawdat does not propose such a transformation. He does not advocate for public ownership of AI infrastructure, for democratic control of technology, or for the redistribution of the wealth that AI creates. Instead, he hopes that individual consciousness will somehow produce collective outcomes that the system is structurally incapable of delivering.

The closest Gawdat comes to a concrete proposal is his Universal Declaration of Global Rights, which he includes as an appendix. This document lists rights that AI should respect, such as the right to privacy, the right to dignity, and the right to self-determination. These are worthy principles, but they are presented as if they can be achieved through moral persuasion alone, without any mechanism for enforcement. The Universal Declaration of Human Rights has existed since 1948, and it has not prevented genocide, exploitation, or inequality. Why would a similar declaration for AI be any more effective? One where the benefits of AI are shared equally. One where technology serves human needs rather than capital. This is a vision that can only be realized if the underlying economic system is transformed. Gawdat does not propose such a transformation. He does not advocate for public ownership of AI infrastructure, for democratic control of technology, or for the redistribution of the wealth that AI creates. Instead, he hopes that individual consciousness will somehow produce collective outcomes that the system is structurally incapable of delivering.

Gawdat's vision of utopia is one where AI solves humanity's problems, relieves us of mundane work, and allows us to focus on connecting and contemplating. It is a vision of a post-scarcity world where technology serves human flourishing. He paints a picture of people sitting around a campfire in 2055, free from the drudgery of labor, able to enjoy nature and human connection. This is an attractive image, but it depends on the assumption that AI will be developed and deployed in a way that serves everyone equally., relieves us of mundane work, and allows us to focus on connecting and contemplating. It is a vision of a post-scarcity world where technology serves human flourishing. This vision is appealing, but it is not achievable within the current economic system.

A world where AI serves human flourishing would require that AI be developed and deployed for the common good, not for private profit. It would require that the benefits of AI be distributed equitably, not captured by the already wealthy. It would require that the decisions about AI development be made democratically, not by corporate executives and venture capitalists. None of these conditions exist, and Gawdat does not propose any realistic path to achieving them.

The utopia that Gawdat imagines is the one that Silicon Valley has been promising for decades: a world where technology solves all problems and everyone lives happily ever after. It is the same promise that was made about the internet, about social media, about smartphones, about every major technological innovation of the past fifty years. And in every case, the promise has not been fulfilled. The technology has been captured by corporate interests, its benefits have been concentrated at the top, and the harms have been borne by the most vulnerable.

There is no reason to believe that AI will be different. In fact, there is every reason to believe that AI will follow the same pattern, only faster and with more devastating consequences. The technology is being developed by the same companies, operating under the same logic, with the same incentives. The only thing that has changed is the scale of the potential harm.

Part Five: The Real Solution

If individual consciousness is not the solution, what is? The answer requires us to think beyond the framework that Gawdat provides. It is not enough to become more conscious as individuals. We must become organized as a class. We must build the political power to take control of AI development away from the corporations that currently own it.

This means supporting efforts to regulate AI development through democratic processes, not corporate self-regulation. It means advocating for public investment in AI research that is not tied to commercial outcomes. It means building and using open-source and federated alternatives to corporate AI platforms. It means supporting workers whose jobs are being automated out of existence. It means demanding that the wealth generated by AI be used to fund public goods, not to enrich shareholders. The answer is collective action to change the system that controls AI development. This means building political power to regulate corporations, to break up monopolies, to establish public ownership of AI infrastructure, and to ensure that the benefits of AI are distributed equitably.

This is a much harder task than becoming more conscious. It requires organizing, building movements, and confronting the most powerful institutions in the world. It requires recognizing that the fight for a just AI is the same as the fight for a just society, and that we cannot have one without the other.

Gawdat dismisses regulation as an effective solution. He argues that regulation moves too slowly, that it can be captured by the companies it is supposed to regulate, and that it cannot keep pace with technological change. There is truth in these criticisms. But his conclusion is not that we need better regulation. It is that we need to rely on individual consciousness instead. This is a false choice. Gawdat's dismissal of regulation is particularly revealing. He writes from within a Silicon Valley culture that has always been hostile to democratic oversight of technology. The tech industry has spent decades fighting regulation, lobbying against privacy laws, and promoting the ideology that innovation requires freedom from government interference. For Gawdat to dismiss regulation while offering only individual consciousness as an alternative is to reproduce the very ideology that created the problem. It means we need to fight for better regulation, for regulation that is designed by and for the people who are affected by AI, not by the companies that profit from it.

Gawdat's book contains a kernel of truth: the future of AI is not predetermined. It will be determined by the choices we make. But the choices that matter are not the choices of individuals to become more conscious. They are the collective choices we make about how to organize our society, who holds power, and what values guide our institutions.

The real question is not whether we will raise AI with the right values, but whether we will build a society where the values that guide AI development are determined by the many rather than the few. Gawdat's book, for all its good intentions, ultimately serves the interests of the very system it claims to critique. By framing the problem as one of individual consciousness, it directs attention away from the structural changes that are actually needed.

This is not an accident. It is a feature of the ideology that Gawdat represents. The tech elite has always preferred individual solutions to structural problems. They prefer meditation over regulation. They prefer consciousness over collective action. They prefer personal transformation over political transformation. This preference serves their interests because it leaves the existing power structures intact. The real question is whether we will build a society where the values that guide AI development are determined by the many rather than the few. And that is a question that cannot be answered by reading a self-help book. It can only be answered by engaging in the difficult work of organizing and building power.

Conclusion

Mo Gawdat's "Scary Smart" is a well-intentioned book written by someone who has seen the inside of the AI industry and is genuinely concerned about where it is heading. His desire to sound an alarm is understandable. His call for awareness is not wrong. But his framework for understanding the problem and his proposed solution are both inadequate.

The problem with AI is not that it has not been raised properly by conscious individuals. The problem is that it is being developed under capitalism, by corporations that are driven by profit, for the benefit of shareholders. The solution is not for each of us to become more conscious. The solution is to change the system that determines how AI is developed and deployed.

Gawdat's book is a symptom of the very problem it claims to solve. It offers individual solutions to structural problems. Gawdat's book is a product of the Silicon Valley ideology that it claims to critique. It believes in technology as salvation. It believes in individual action as the driver of change. It believes that the problems created by capitalism can be solved within capitalism. These beliefs are not true. They are the ideology of the class that benefits from the current system. Gawdat's book is a symptom of the very problem it claims to solve. It offers individual solutions to structural problems. It asks us to look inward when we should be looking outward at the systems of power that shape our world. It tells us that we can save the world by changing ourselves, when the truth is that we can only save the world by changing the system. The future of AI will be determined not by how many individuals become conscious, but by whether we can organize to take control of the technology away from the corporations that own it and place it under democratic control.

A real movement for AI justice would be built by workers, not by tech executives. It would be led by the people who are most affected by AI: the warehouse workers being monitored by algorithms, the call center workers being replaced by chatbots, the artists whose work is being used to train models without consent, the communities being targeted by predictive policing systems. These are the people who have the most to gain from changing the system and the most to lose from leaving it unchanged. It would demand a moratorium on the deployment of AI systems that cannot be explained or challenged by the people they affect. It would demand that workers have a say in how AI is deployed in their workplaces. It would demand that the data used to train AI systems be treated as a public resource, not a private asset to be extracted for profit. It would demand that AI research be funded publicly and that its results be made openly available, not locked behind corporate firewalls. He asks them to commit to being part of the solution, to spread the message, to become conscious of the AI challenge. These are not bad things to ask. But they are not enough. A movement that does not challenge the power of the corporations that control AI is not a movement that can save us. A movement based on individual consciousness does not threaten the power of the tech giants. It does not challenge their control over AI development. It does not demand that they share the benefits of AI with the people whose labor and data made it possible. It asks only that those individuals who are worried about AI become more aware. This is not a movement. It is a coping mechanism.

The future of AI will not be determined by how conscious we are as individuals. It will be determined by whether we can organize to take control of the technology away from the corporations that currently own it and place it under democratic control. That is the real task. And it is a task that requires not consciousness, but collective action.

📺 Watch on YouTube: https://youtu.be/HmXTFJYZ7CY
🌐 Watch on PeerTube: https://tankie.tube/w/f9c2676a-dd5c-4d48-b990-eea135512abf

📡 I'm live on Twitch: https://www.twitch.tv/jorvex609
📱 All links: https://linktr.ee/jorvex609