Opinion: If the current AI boom is in fact a bubble, I'm willing to bet that it won’t be the tech bros who pay the price
I recently rewatched 'The Big Short'. As an Oscar winning movie with a stellar cast and brilliant performances, especially from Steve Carrell and Christian Bale, it’s worth a watch in and of itself. But if you’d like to understand what caused the 2008 global financial crisis, it’s a must-watch. The movie is dubbed a comedy, but outside of Margot Robbie explaining subprime mortgage bonds whilst sipping champagne in a bathtub, this time I didn’t find much to laugh about.
What struck me rewatching it was how little we have learned since the crash. In the years leading up to it, banks created property market investment vehicles so complex, few people understood the underlying high levels of risk. Bullish investors excessively speculated billions, feeding a bubble that eventually burst and took down the world economy. I couldn’t help but see parallels with the current AI boom; huge levels of investment in technologies that most users don’t fully understand, and investors downplaying the risky nature of forecast future returns.
From Paramount, The Big Short official trailer
There is every chance that the AI boom is another bubble, but the problem is that in the midst of a bubble, it can be hard to see the wood for the trees. In a boom, investment increases because of strong fundamentals; an increase in the value of a product as defined by higher benefits or returns. In a bubble, investment growth is driven by over exuberance and optimism, going beyond the levels justified by the underlying value of the asset. While there are clear benefits to AI in some areas and applications, the scale of investment in developing generative AI and the infrastructure required to support it is staggering. Between now and 2029, global spending is set to hit $3 trillion, an amount equivalent to about 15% of the European Union's GDP. It’s hard to see how the productivity gains from AI, which in many domains are still only forecast rather than realised, can justify such huge costs.
Although the use of generative AI undoubtably can increase output, it can often be a question of quantity at the expense of quality. AI can speed up certain tasks, but its use in many situations is akin to asking chefs to switch to using microwaves to cook food. Yes, you can produce more meals more quickly, but it’s not going to taste as good as it would if you used other tools. Error rates in information provided by chatbots continue to be high, with peer reviewed research finding, for example, that 66% of answers related to drug information for patients provided by bots is potentially harmful, and bot reviews of existing literature on a topic miss up to 96% of relevant studies.
From PBS News Hour, Is a bubble forming as AI investments drive economic growth?
Even if future advances in the technology eliminate these errors and lead to the promised productivity gains, the costs of development are not confined to the monetary costs of development and supporting infrastructure. The environmental costs of the increased energy demands of AI are staggering. According to the International Energy Agency, global emissions from data centres has almost doubled in the past five years and will continue to increase up to 2030 with 50% of data centre power generated by burning fossil fuels. In Ireland over the past five years, the rise in energy used by data centres has far outpaced growth in in renewable energy generation, eroding any progress made. Current projections are that we will not meet our legally binding emissions reduction targets, and face compliance costs in the form of fines to the EU of up to €26 billion.
There's a scene at the end of The Big Short where Mark Baum (a fictionalised version of Steve Eisman, one of the few investors who bet against the US housing market bubble) learns that the banks who invested heavily and lost, are getting a bailout. He comes to the realisation that 'They knew, they knew the taxpayers would bail them out. They weren’t being stupid, they just didn’t care.’ Many banks engaged in high risk speculation, safe in the knowledge that they were too big to fail.
We need your consent to load this rte-player contentWe use rte-player to manage extra content that can set cookies on your device and collect data about your activity. Please review their details and accept them to load the content.Manage Preferences
From RTÉ Radio 1's Today with Claire Byrne, Dr Andrew Hines, lecturer at the School of Computer Science, UCD on how AI is impacting your life right now
Again, there are parallels with the current AI boom; four companies, Meta, Microsoft, Google, and Amazon will spend a projected $750 billion on data centres between this year and next. To put that in perspective, that equates to 130% of Irish GDP in 2024. These companies, betting big on AI, are also huge marketing machines, hyping up its potential. They provide and support platforms critical for the internal and external communications of government agencies and institutions worldwide, and for billions of businesses and private users internationally. Given the dominance of their applications across the globe, it’s likely that they’re not blind to the risks of AI investments, rather they know that like the banks, they’re too big to fail.
There is widespread acknowledgment that the current AI boom is propping up the US economy, masking some of the negative effects of Trump’s chaotic, erratic economic policies. But what happens if the promised advances are not realised and the bubble bursts? To quote Mark Baum again: "I have a feeling that in a few years, people are going to be doing what they always do when the economy tanks, they will be blaming immigrants and poor people". If the current AI boom is in fact a bubble, I’m willing to bet that it won’t be the tech bros who pay the price.
Follow RTÉ Brainstorm on WhatsApp and Instagram for more stories and updates
The views expressed here are those of the author and do not represent or reflect the views of RTÉ
