The Dawn of a New Era in Technological Innovation

The chart below depicts a remarkable surge in Nvidia’s quarterly revenues, which have soared to about $25 billion—over a five-fold increase since early 2024.

Major tech companies like Google and Microsoft are observing these numbers in disbelief. Analysts who think Nvidia has peaked may need to reevaluate their forecasts.

“Commentators in the mainstream often fall victim to the idea that ‘it’s just about predicting the next word.’ They are caught up in hype and see a mere continuation of business as usual; at most, they entertain the notion of another internet-scale tech transformation,” comments Leopold Aschenbrenner in his collection of essays titled The Decade Ahead.

Before long, a revelation is forthcoming. What’s about to unfold is both exhilarating and potentially concerning.

Artificial ‘superintelligence’

AI is progressing towards AGI (Artificial General Intelligence) at a much faster rate than initially expected, allowing machines to surpass human cognitive abilities. This leads to ASI, or artificial superintelligence, on the horizon.

Nvidia is not the only player in the AI chip landscape. Google has launched its Tensor Processing Units to optimize machine learning tasks.

Nvidia dominates the AI chip market with over 70%, while competitors like Intel and Advanced Micro Devices are eager to nibble away at this vast market share.

There’s no denying Nvidia’s significant head start—often referred to as a moat—but its attractive gross margin of 78% is likely to entice intense competition, which seems imminent.

The AI chip sector is anticipated to achieve $400 billion in annual revenue by the end of the decade, paving the way for many newcomers with specialized chips designed for specific functions.

Read/listen:
WEF forecasts rapid growth in AI, big data, and cybersecurity jobs
Nvidia and its associates bolster SA unit trusts
Magic and Warhammer: The influence of wide-moat IP

The cost of AI is decreasing at a rate faster than any previous disruptive technology.

“The cost to run artificial intelligence models of equivalent performance has been halving every four months—a trend we expect to persist throughout this decade. In contrast, Moore’s Law in the semiconductor industry has historically halved costs every 18-24 months, suggesting that the AI revolution is advancing 4-6 times faster,” asserts ARK Invest.

Rumors suggest Microsoft and OpenAI are partnering on a $100 billion data center focused on an AI supercomputer called Stargate, set to launch in 2028.

Considerable investments are being directed towards similar projects across the US, China, Dubai, and various other regions. Microsoft and Google are poised to spend over $50 billion each in capital expenditures in 2024, with Meta closely following. Although not all this spending is allocated to AI, the trend is unmistakably evident.

Source: Best Investing

According to Leopold Aschenbrenner, companies are more inclined to invest heavily in AI if they believe the economic returns will justify their investments.

OpenAI has reported a revenue doubling to $2 billion in the six months ending February 2024. If this trend continues every six months, it could reach $10 billion by early 2025.

Microsoft is projected to earn an additional $5 billion in revenue from AI.

This suggests that substantial investments in AI are likely to yield rapid returns. As organizations like Microsoft, Google, and Meta start to derive a significant share of their revenues from AI—an occurrence anticipated in just a few years—the demand for capital will undoubtedly rise.

If Microsoft can convince a large portion of its 350 million paid subscribers to pay an extra $100 a month for an AI upgrade, revenue growth could spike considerably.

“The implications are monumental. This would position AI products as the core revenue source for the largest corporations in the United States and signify their most crucial growth avenue. Overall revenue growth projections for these companies would surge,” adds Aschenbrenner.

“The stock markets would respond accordingly; we might soon witness the emergence of our first $10 trillion company. At that point, big tech would likely be compelled to invest heavily, with each pouring several hundreds of billions into further AI development. We could potentially see our first corporate bond issuance exceeding a hundred billion dollars then.”

The power constraint

The primary limitation in the AI race lies with power supply. Power generation in the US has increased by a mere 5% over the last ten years; however, the proposed establishment of a 100-gigawatt AI cluster would consume about 20% of the total US electricity supply on its own.

To meet the escalating demands of AI, tech companies may need to enter the power generation sector or acquire aluminum smelters for their substantial power contracts.

While clean energy would be the preference, a Trump administration is more prone to advocate for US dominance in the AI field to prevent China or a Middle Eastern nation from achieving a lead. This might involve leveraging abundant natural gas reserves and reducing green energy initiatives.

The ramifications of lagging in AI, on a governmental level, are staggering.

Superintelligence could be weaponized. Rogue nations might use it to issue threats of annihilation or to sabotage adversarial states, potentially neutralizing traditional nuclear deterrents. Countries that secure even a slight advantage in AGI will hold a significant competitive edge.

Aschenbrenner contends that the US must secure this race or risk losing AGI supremacy to authoritarian governments.

“These clusters can be formed in the US, and we must unite our efforts to ensure this takes place domestically. American national security must take precedence over the lure of unregulated capital influx from the Middle East, complicated regulations, or even commendable climate pledges.”

South Africa is falling behind in this race. The significant energy demands of this burgeoning sector cannot be met in a country that has struggled to maintain a stable electricity supply for much of the past decade, with the past year being a rare exception.

Stay informed with Moneyweb’s extensive finance and business news on WhatsApp here.