The biggest technology firms in the world posted almost universally disappointing second-quarter earnings for 2024. Though there were a couple of exceptions, companies that have attached their wagons to the generative artificial intelligence train appear to be losing ground with mainstream consumers and enterprise clients alike.
All indications point to the ouroboros problem — often depicted as a snake eating itself — where generative AI tools such as chatbots can only become so useful before increasing the amount of data and processing power they use produces diminishing returns.
The same appears to be true when it comes to speculative spending on AI technologies. If Q2 2024 is any indication, the global AI sector may have problems that money can solve.
Chatbots aren’t enough
When OpenAI launched ChatGPT in November 2022, the technology world seemed to let out a collective gasp. At the time, based on how venture capitalists and Big Tech budgets responded, it seemed like anything was possible in the world of generative artificial intelligence.
Now, we’ve come full circle, as OpenAI is reportedly facing operating losses without another cash injection, despite having launched GPT-4 and GPT-4o in the interim.
In fact, Google just launched the world’s smartest chatbot according to Chatbot Arena benchmarks. It beat the most advanced ChatGPT and Claude models to take the top leaderboard slot in overall performance. And seemingly, the general public didn’t even notice.
Part of the problem is that firms such as OpenAI, xAI and Anthropic have all stated that their sole mission is to develop artificial general intelligence (AGI). This has led to hundreds of billions of dollars of investments and spending on generative AI technologies throughout the sector.
But there’s no actual scientific consensus as to what an AGI is or whether it’s even fundamentally possible. And without AGI, there may not be enough runaway for continued investment.
Artificial general intelligence
AGI is a nebulous term for a theoretical AI capable of performing any feat of reasoning or intelligence that a human could, given sufficient resources.
Those companies that solely exist for the purpose of creating AGI will ultimately have to create products and services that Fortune 500 companies rely on, find a way to monetize consumer use (such as advertising), or actually invent AGI.
Right now, the majority of potential enterprise clients exploring generative AI products and services are doing so under the assumption that they’ll need to be prepared for an AI-powered future. However, there’s little evidence that services such as Grok or ChatGPT have become integral to enterprise operations in the same way that, for example, machine learning algorithms have.
It’s difficult to measure consumer interest in generative AI technologies because little of the sector’s profits come directly from consumers. Most of the major chatbot players offer subscription services, but data shows that subscriber numbers for gen AI are nowhere near those of streaming entertainment services. Without a killer app for business or consumers, generative AI technology evidently remains more of a curiosity to both markets as we embark on the downhill half of 2024.
Flat spin
The other companies involved in the generative AI space — including Microsoft, Google, Apple, Meta, Nvidia, Amazon and Tesla — are also beholden to the promise that AGI is both possible and imminent.
The difference between them and the dedicated firms, such as OpenAI and Anthropic, is that they’ll eventually pivot away from generative AI if it doesn’t pan out and another avenue toward AGI doesn’t present itself.
Amazon, for example, just took a 9% hit to its stock after CEO Andy Jassy indicated the company would prioritize bringing AI products/services to market over profits. This signals a rare disconnect between big tech executives and shareholders when it comes to betting big on generative AI in the wake of ChatGPT’s launch.
The big question going forward is whether the AI sector can pull out of the flat spin of spending to break even or if Big Tech firms are going to hit the eject button on chatbots.
Based on Q2’s earnings and the subsequent market movement — only Apple and Tesla remain up over the past 30 days — it may be shareholders and investors who ultimately pull the plug. Unless, of course, someone actually manages to bring an AGI model to market.
Related: Decentralized AI gains ground as more protocols collaborate