Another week brings another all-time high in US equity markets, with last week’s surge attributed to the Federal Reserve’s indication of potential further interest rate cuts in the coming months. However, the underlying optimism in the market is underpinned by two primary factors: the substantial cash reserves held by tech behemoths now dominating the markets and the belief in their capacity to capitalize on artificial intelligence.
The pervasive narrative extols the transformative power of AI, promising a significant boost in productivity (albeit at the expense of displacing numerous jobs) and the creation of a vast new wealth pool for global sharing. An ARK Invest report, released last week, prophesies a staggering $40 trillion surge in global GDP by 2030 due to AI, asserting its pervasive influence across all sectors, businesses, and innovation platforms.
Yet, it is the exuberance and air of inevitability surrounding this narrative that evoke a sense of caution. Despite the comparison of AI to past revolutionary technologies like electricity and the internet, we are merely at the nascent stages of a highly intricate, multi-decade transformation that is far from a foregone conclusion. Nonetheless, current valuations seem to encompass the entire anticipated paradigm shift and more. A report from Currency Research Associates in February highlighted that it would take 4,500 years for Nvidia’s future dividends to equal its present price—a truly extensive horizon.
While Nvidia is not Pets.com—due to its tangible revenue streams from actual product sales—the broader AI storyline hinges on numerous uncertain conjectures. For instance, the substantial water and energy demands of AI necessitate scrutiny, with both the US and EU advocating for enhanced transparency regarding companies’ resource consumption. It is plausible that these input costs will witness a significant escalation in the future, whether through carbon pricing mechanisms or resource usage taxes.
Furthermore, AI developers are not mandated to possess copyright ownership of the data used to train their models. Profit generation directly from AI operations is not imperative; the mere anticipation of future returns propels the fervor. Silicon Valley thrives on unwavering techno-optimism and the illusion of inevitability to generate substantial paper wealth. However, it is crucial to remember that advocates of “AI everywhere” were recently championing concepts like web3, cryptocurrencies, the metaverse, and the gig economy.
A notable distinction lies in the validation of AI by major cash-rich industry leaders such as Microsoft, Google, and Amazon. Nevertheless, skepticism persists even within these firms, with a senior executive at a prominent AI company conceding that profit assumptions concerning the technology are more speculative than substantive, with significant challenges yet to be addressed.
Anyone familiar with large language models can attest to this reality. Relying on a chatbot for research purposes may raise concerns about data accuracy and the relinquishment of control over informational inputs. Even for more routine tasks across various job segments, integrating AI into workflows raises pertinent questions about its efficacy compared to human counterparts. The resistance is palpable, with Hollywood writers’ strikes and union activism reflecting concerns over AI control and technology regulation.
The mounting backlash against AI’s copyright infringements is gaining momentum, exemplified by French regulators fining Google €250 million for unauthorized use of news articles to train its AI algorithms, alongside similar legal actions against OpenAI and Microsoft by the New York Times. As AI permeates proprietary corporate datasets, the potential for copyright-related litigations is poised to increase, possibly intersecting with grievances over corporate surveillance practices.
Moreover, the issue of monopolistic control looms large. According to Meredith Whittaker, president of the Signal Foundation and co-founder of the AI Now Institute, contemporary AI advancements are predominantly driven by a handful of tech giants wielding substantial data and computing resources. This escalating dependence on such AI resources cedes disproportionate influence over individuals and institutions to a select group of tech corporations.
The surge in AI enthusiasm and stock market gains over the past year has been spearheaded by the so-called Magnificent Seven companies, propelling the concentration of the S&P 500 to unprecedented levels. Nonetheless, a recent report from Morgan Stanley Wealth Management underscores that historical trends indicate self-correcting mechanisms in index concentration, with regulatory, market, and competitive dynamics, alongside business cycle fluctuations, undermining static leadership positions. The report suggests that equity returns typically encounter challenges following peaks in concentration.
This corrective interplay may encompass a myriad of factors, including the proliferation of Big Tech antitrust litigations and the potential impact of carbon pricing and copyright penalties on the “free” inputs crucial for profit generation.
Whether AI is perceived as the next speculative bubble or a groundbreaking innovation akin to the combustion engine, it is imperative to scrutinize how the market is valuing this narrative.
