Innovation with grains of salt

The early 21st century is obsessed with innovation. Google’s Ngram shows that the term “innovation” appears nearly 2.5 times more frequently than it did 20 years ago. Searching the Web of Science, which indexes nearly 10,000 periodicals, now returns more results for “innovation” than for “economy.” Innovation has become a new mantra, more than just a universal promise of relentless progress but increasingly a casually bestowed label for all kinds of sweeping AI-driven planetary transformations. I urge deep scepticism or at least abundant caution.

In medicine, recent promises have included not only groundbreaking advances in the repeatedly invoked and relaunched war on cancer and in delaying and beating age-related cognitive decline, but also the arrival of individualised medical care based on a deep understanding of a person’s genetic make-up. While there have been welcome and important treatment gains, such as the rapid deployment of new mRNA vaccines to control the Covid-19 pandemic and rising survival rates for head and neck cancers, we have not seen any sweeping, grandiose progress. We must also reckon with some notable retreats in overall health outcomes, including average life expectancy and the availability of basic medical care.

By 2023, life expectancy in the US had declined to its lowest level in two decades. In some affluent countries, including Canada, Australia, and Norway, patients (many in pain) are waiting longer for common surgeries such as hip and knee replacements. Europe’s ageing population is facing a shortage of 1.8 million health workers, and no country with an ageing population, with China currently topping the list in absolute terms, is prepared to cope with the rising tide of physical and mental care that will be required in the near future. None of these challenges will be solved by the mass consumption of anti-obesity drugs.

As for the state of the global environment, we often hear about how high-tech advances—from offshore wind mega-farms to cheap green hydrogen, and from high-gain heat pumps to the electrification of everything—will decarbonise global energy use by 2050. However, since the Kyoto Protocol set the first goals for reducing global CO2 emissions back in 1997, emissions have risen by 61 percent. In 2023, they reached yet another record level (see Figure 1). The chances of reversing this trend suddenly are vanishingly small. To reach net zero by 2050, the drop would have to average about 1.5 gigatonnes (Gt) CO2 per year, an annual cut equivalent to the combined 2023 emissions of Germany, France, Italy, and Poland.

What is more, available technical fixes face two daunting challenges: mass scaling and affordable costs. The world will produce some 0.4 megatonnes (Mt) of green hydrogen this year compared to about 95 Mt of black hydrogen derived from hydrocarbons. Decarbonising the global primary steel and synthetic ammonia production projected for 2050 alone would require nearly 150 Mt of green hydrogen. Producing this amount by water electrolysis would require scaling its 2024 output nearly 400-fold, with even more green hydrogen needed for other hard-to-electrify decarbonisation uses, especially in industrial production and transportation.

We are also not moving away from but rather adding to new fossil fuel converters: more than a thousand new jetliners (requiring aviation kerosene), hundreds of giant container ships (burning diesel fuel and heavy fuel oil), and dozens of new large blast furnaces (charged with coke, coal dust, and natural gas) and rotary cement kilns (running on any low-quality heavy fuel) are being added each year. None of these have any non-carbon commercial alternatives that would be readily available at scale and widely affordable. And while installing a heat pump (at CHF30,000 or USD34,000) may seem like a (subsidised) bargain to wealthy Westerners, that sum is about 10 times the annual average income in Nigeria, making this option entirely irrelevant for billions of people in low-income subtropical and tropical climates where air conditioning is in the highest demand.

Both health- and energy-related innovation hypes have been cautious compared to the exaggerated claims made on behalf of artificial intelligence—especially those made over the past 25 years by Ray Kurzweil, an American computer scientist, inventor, and a former director of engineering at Google. In 2005, he claimed that the singularity, as in when “machine intelligence will surpass human intelligence, leading to The Singularity—technological change so rapid and profound it represents a rupture in the fabric of human history” is near. His latest book published in June 2024 argues it is “nearer” (no later than 2045), predicting that the merger of biological and nonbiological intelligence will produce immortal software-based humans “and ultra-high levels of intelligence that expand outward in the universe at the speed of light.”

Kurzweil believes that when AI “gives us full mastery over cellular biology,” the annual increase in life expectancy will greatly accelerate and that “for people diligent about healthy habits and using new therapies, I believe this will happen between 2029 and 2035—at which point ageing will not increase their annual chance of dying.” So, just hang on until 2029, and eternal life on an AI-ruled Earth is yours! Admittedly, most of the recent media-propagated AI logorrhoea, as overwrought as it has been, does not go as far as Kurzweil’s irrepressible claims, but it is still firmly in the realm of supernaïve hype.

When Marc Andreessen, a general partner of a leading US venture capital firm, says that “AI will save the world” and that it “can make everything we care about better,” does he mean that, or does this hyperbolic claim apply only to information management? If the former, then I urge readers to make their own short lists of such “save” and “care” measures and ask what AI will do for them in five or 10 years. My list, with inclusions guided by their overall potential to save lives, would include the complete elimination of nuclear weapons, the economic development of Africa, and the end of malnutrition. I do not see the vaunted large language models (LLMs) and generative AI (Gen AI) triggering fundamental transformations in society, crime, or politics. And if not, what then is that “everything we care about” which AI will do for us? Writing personalised rejection letters or drawing cartoons in Picasso’s style?

The last 150 years have been an age of unprecedented scientific advances and technical inventions, but a critical look reveals that their pace has not been generally accelerating. Computing capabilities have been the single most important exception as we moved from vacuum tubes to solid-state electronics (by the mid-1950s), then to integrated circuits (starting in the late 1950s) and microprocessors (starting in 1971), whose soaring performance enabled the extremes of supercomputing and personal portable devices. But the physical (truly existential) fundamentals of modern civilisation have not seen any similarly radical changes over the past 50 years.

We generate most of our electricity using large steam turbogenerators, make primary steel in blast and basic oxygen furnaces, process cement in rotary kilns, power intercontinental flights with gas turbines, and run rapid trains with electric motors and large ships with massive diesel engines. We achieve high crop yields by applying synthetic fertilisers and keep finding new uses for hydrocarbon-derived plastics. The world of 2024 is just a version of the world of 1974. What has changed in all these cases are the incremental gains in efficiency and performance, and hence the cost and environmental burden of these fundamental activities. These have been the most consequential innovations. Thanks to them, we need less energy per unit of product or service, making these more affordable and more acceptable, and they have improved many lives by providing better nutrition, better housing, and higher incomes. Keep this in mind when coming across yet another uncritical claim about the latest transformative, epoch-making, world-changing innovations.