Big Tech companies like Apple and Alphabet, Google's parent company, are anticipated to utilize artificial intelligence (AI) to sustain their industry leadership, as outlined in a white paper released by Cathie Wood’s investment firm ARK Invest on October 1. The document, authored by ARK’s chief futurist Brett Winton, delves into the disruptive potential of AI and presents a framework for identifying such technologies.
Winton's report highlights how major technology firms are likely to enhance their AI capabilities to maintain their market dominance while also acknowledging potential pitfalls in these strategies. He identifies three core properties of disruptive technology platforms: significant cost reductions, penetration into new or underserved markets, and business models that delay monetization and may initially appear financially unattractive.
These characteristics often enable smaller firms to compete with industry giants, even when the latter recognize the technology’s potential and attempt to leverage it for business gains. Winton emphasizes that AI is a highly disruptive technology, noting its remarkable cost decline; the cost to operate AI models with equivalent performance has been halving every four months, a trend expected to continue throughout the decade.
This rapid decline in costs is compared to Moore’s Law in the semiconductor industry, which suggests that semiconductor costs halve every 18 to 24 months. According to Winton, the AI revolution is progressing four to six times faster than historical technological advancements. The report also elucidates how incumbent tech companies often allow startups to de-risk new technologies before adopting them on a larger scale.
Winton points out that Google and Apple have adhered to this strategy with AI. For instance, Google refrained from publicly releasing a large language model until OpenAI had been in the market for over three years. Despite significant marketing efforts, Google’s performance lagged behind OpenAI, with the most advanced Google model costing customers over 40% more in per-unit performance compared to OpenAI’s model.
While Google has trailed OpenAI, Apple has yet to launch a large language model. Winton anticipates that Apple will introduce its first modern AI-driven products in the fall of 2024. He notes that a slower approach does not necessarily indicate a competitor will lose the race, as companies like Google and Apple have valid reasons for delaying AI features.
A primary concern is the risk associated with releasing a product that exhibits unpredictable performance, which poses significant challenges for companies with meticulously cultivated reputations. However, Winton raises a critical question about whether this cautious strategy will empower Big Tech firms to roll out widely effective AI systems, concluding that incumbents may prefer less disruptive technologies, potentially putting them at a disadvantage..