The AI industry is facing an impending crisis, and it's time to salvage what we can from the wreckage. As a science-fiction writer, I'm here to interrogate the techno-social implications of AI, not predict the future. But let's be clear: AI companies will fail, and it's crucial to understand why.
The delusion that science-fiction writers can foresee the future is widespread, even among my colleagues. This misconception has led to a depressing trend of AI enthusiasts, or 'AI bros', who believe they are witnessing the future. Their obsession with AI's potential has led to a surge in investment, with big investors pouring hundreds of billions of dollars into AI companies. But here's the controversial part: these investments are based on a flawed narrative.
The growth story of AI revolves around disrupting labor markets, promising to replace human workers with AI. This narrative is a bait-and-switch. AI can't do your job; it can only assist you. Yet, investors are betting on AI replacing radiologists, for example, and saving companies money. But the reality is, AI can't do a radiologist's job, and firing them would only make radiology more expensive and potentially less accurate.
AI companies are creating 'reverse centaurs'—humans serving as appendages to machines. Take Amazon delivery drivers, monitored by AI cameras, penalized for looking away or singing, and pressured to meet superhuman quotas. These drivers are not benefiting from AI; they are its victims.
Tech bosses want us to believe there's only one way to use technology, but we must challenge this. Mark Zuckerberg, Tim Cook, and Sundar Pichai want us to think their platforms are indispensable, but they're not. This is Thatcherism in action: 'There is no alternative.' But we must reject this fatalism.
The AI bubble is fueled by monopolies like Google and Meta, who control markets and stifle competition. These growth-stock companies are desperate to keep investors convinced of their potential, creating one bubble after another. But the paradox is, once a company dominates, the market slashes its value if it doesn't trust its pricing power.
AI art is another controversial aspect. AI-generated art is often used as an advertisement, but it's not part of AI's business model. The purpose is to create buzz and convince the public that AI is amazing. But AI art is bad art; it's eerie, lacking the creative intent of human artists. Expanding copyright to cover AI training won't help artists; it will only benefit the companies that control creative labor markets.
The US Copyright Office has made a crucial decision: AI-generated works cannot be copyrighted. This means AI-created works are in the public domain, and anyone can use them. This is a recipe for 'centaurhood,' where artists and writers can use AI tools without losing their rights. But we must also organize and demand sectoral bargaining, not just expanded copyright.
AI is a bubble, and it will burst. Most companies will fail, and datacenters will be shuttered. But we can salvage useful AI tools that run on our laptops and phones, doing things like transcribing audio, describing images, and automating graphic editing. These tools are the silver lining in the AI wreckage.
So, let's be clear: AI is not the future. It's a bubble, and we must fight to protect workers and the public from its negative impacts. Let's burst the bubble and reclaim our technological society from the grip of monopolies and unethical investors.
