Can AI Sustain Its Market Momentum?

AI's Ethical Challenges and Future Developments

by Faruk Imamovic
Can AI Sustain Its Market Momentum?
© Getty Images/Justin Sullivan

The stock market is no stranger to hype-driven surges, and the latest wave centers on artificial intelligence (AI). Giants like Nvidia, Microsoft, Apple, and Alphabet have collectively seen their values soar by $1.4 trillion in just the past month. Remarkably, Nvidia alone accounted for half of this staggering gain. However, this meteoric rise prompts a crucial question: Are we witnessing the birth of another dot-com-style bubble, or is there substantial merit behind these valuations?

The Rise and Risk of AI Stocks

Unlike the speculative frenzy of the dot-com era, Nvidia's profitability is currently rising in tandem with its stock price. This suggests that the company's value isn't solely driven by hype. However, the sustainability of these profits is uncertain. As AI technologies proliferate, the initial excitement might give way to skepticism if consumers start feeling that AI is overhyped and fails to meet expectations. This potential shift in sentiment could precipitate a swift decline in earnings and stock prices.

Paul Romer, a Nobel Prize-winning economist, draws a parallel between the current AI enthusiasm and the recent crypto boom. "There was this solid consensus only a couple of years ago that cryptocurrencies were going to change everything, and then suddenly that consensus just goes away," Romer notes, cautioning that investors might be overestimating the future growth of AI. He predicts a significant slowdown, labeling the current scenario as "typical bubble hype where people are trying to cash in on the latest trend."

Exponential Growth in AI Compute Power

Adding another layer to the AI narrative is the exponential increase in computing power, or "compute," required to train high-end AI models. According to a report by Epoch AI, the compute needs for major language models are growing by 4x to 5x annually. Between June 2017 and today, the requirements have surged by up to 9x annually. This explosive growth in compute power underscores the rapid advancement of AI technologies but also highlights potential challenges in scaling and resource allocation.

ChatGPT, a leading language model, estimates that AI compute could account for 1% to 2% of the world's total computing resources. If current trends continue, AI could theoretically consume 4% of global compute next year, 16% the year after, 64% in 2027, and an unsustainable 265% by 2028. These projections, while illustrative, emphasize the immense strain AI development could place on global computing infrastructure.

Exponential Growth in AI Compute Power
Exponential Growth in AI Compute Power© Getty Images/Leon Neal

AI in Political Campaigning and Stock Predictions

The influence of AI extends beyond stock markets and computing resources. In India, politicians have leveraged AI to create deep fake versions of themselves, making more than 50 million voice-cloned political calls in just two months. This strategy allows politicians to reach voters on a personal level without physical presence. However, the ethical implications are significant, as the use of AI is not disclosed to voters, leading many to believe they are having genuine one-on-one conversations with their representatives.

AI's potential in financial markets is equally intriguing. Research from the University of Chicago School of Business suggests that GPT-4, an advanced language model, can predict company earnings and select stocks more effectively than human analysts. By analyzing company financial statements and using "chain of thought" prompts to mimic human reasoning, GPT-4 achieved an accuracy rate close to 60%. Trading strategies based on these predictions have reportedly yielded higher returns compared to traditional models, indicating a promising, albeit nascent, role for AI in financial decision-making.

Advancements in Medical and Legal AI Applications

AI's transformative potential extends to healthcare and legal domains as well. Caristo Diagnostics' CaRi-Heart AI technology, recently featured in The Lancet, has demonstrated the ability to predict fatal and non-fatal heart attacks up to a decade in advance. The study, involving 40,000 patients, revealed that CaRi-Heart could identify risks even in patients with minimal coronary plaque, showcasing AI's capability to revolutionize preventive healthcare.

In the legal arena, a bipartisan group of U.S. senators is advocating for the NO-FAKES Act, aimed at preventing unauthorized AI-generated replicas of actors' and musicians' likenesses. This legislation underscores the need to balance technological innovation with ethical considerations and individual rights. As Senator Amy Klobuchar stated, "There’s going to be great innovations with AI, but we have to allow people to own their own voices and their own artistic works."

The Sentience Debate and Future of AI

The rapid advancement of AI technologies has also reignited philosophical debates about AI sentience. A survey by the Sentience Institute found that 20% of U.S. adults believe some AIs are already sentient, with 38% favoring legal rights for AI and 71% asserting that AIs deserve respectful treatment. Despite these perceptions, experts like the cofounders of the Institute for Human-Centered Artificial Intelligence argue that large language models (LLMs) lack subjective experiences and merely generate coherent-sounding outputs without understanding their meaning.

Ben Goertzel, founder of, is actively working towards creating genuinely sentient machines but believes that LLMs are not the solution. He envisions a future where smarter AI systems, integrating symbolic reasoning and evolutionary learning, will offer better reasoning and creativity. Goertzel's efforts, alongside the scaling of decentralized AI infrastructures, highlight the ongoing quest to develop advanced AI while addressing its current limitations.

The current surge in AI stocks reflects both genuine advancements and speculative enthusiasm. While companies like Nvidia are seeing real profitability, the long-term sustainability of these gains remains uncertain. As AI continues to evolve, its impact will be felt across various sectors, from finance and healthcare to politics and law. However, the ethical and practical challenges accompanying AI advancements necessitate careful consideration and regulation to ensure beneficial outcomes for society.