Technology has long been a measure of human advancement and power. Today, the next revolution is well underway, and the world is rushing to adapt to the rising speed of development. What feels like incremental change today is truly exponential compared with the past, and expectations continue to rise. Disruption is the new normal and will drive relentless change.
AI growing up. Artificial intelligence is reliant upon three major factors: data, computing power and algorithms. The last few years have seen an explosion in computer power and a dramatic increase in the sophistication of algorithms. From predictive maintenance schemes to consumer-facing platforms like Apple’s Siri and Amazon’s Alexa, AI momentum has begun to reach a palpable tipping point. The time to train a large image classification system on cloud infrastructure has fallen from three hours in 2017 to 88 seconds in 2019, with similar decreases in cost. Computing power is now doubling every 3.4 months (significantly faster than the two years of Moore’s Law). Major scientific AI barriers still remain – preventing the fulfillment of many predictions about AI’s capabilities – but leave the door open for innovators to invest and unlock the true potential of AI.
Personalized medicine commercializes. The completion of the Human Genome Project and the development of new techniques like CRISPR and CAR T-cell have revolutionized genomic editing. While personalized medicine at scale continues to be a dream, these techniques have made the development of niche, focused therapeutics possible. Forty-two percent of molecular entities approved by the FDA were personalized treatments, including the second approved cancer treatment based on biomarkers rather than tumor type.89 A string of start-ups has emerged over the last 10 years to bring this technology to life, while large pharmaceutical companies rush to play. Gilead, for example, bought Kite Pharma, a CAR T-cell-based therapeutics producer, for $11.9 billion in 2017.90
Battery development continues. The promise of a zero-carbon grid relies upon the costeffectiveness of energy storage. As renewable infrastructure and EV adoption ramp up, battery development has exploded in the last five years. Tesla’s partnership with Panasonic galvanized the industry with the Gigafactory, the world’s largest battery production plant, at a planned 35 GWh capacity. The two have sparred about production matching promised yield, but the growing pains seem to be over as the plant has hit stated capacity and more is being planned.91 VC investment into battery development increased over 100% in both 2018 and 2019 as start-ups emerged to commercialize the next phase of batteries from solid state to flow. 92
Quantum computing emerges. While development will continue for the next five years, quantum computing has made rapid advancements over the last few years, with IBM and Google leading the industrial effort. In 2019, Google announced that their Sycamore computer had established "quantum supremacy" by completing a task that would be nearly impossible with a traditional computer.93 In a surprise announcement in 2020, Honeywell announced that they built the world's best quantum computer and planned to launch its capabilities to client via the internet in 2020.94
Michael Rogers, former futurist at The New York Times, has applied a sailing analogy to the process of innovation: know your direction and tack the sail, bit by bit, against headwinds to reach the destination. Science and technology over the next five years will continue tacking, culminating in breakthrough moments and the beginnings of mass arrival for many technologies. The biggest opportunities for companies are in faciliating that arrival and readying themselves before it comes.
Companies that chip away at major obstacles to breakthrough technologies are likely to reap significant rewards. For example, personalized medicine's current high cost and inaccessibility are largely because scaling manufacturing is very difficult. Next-generation approaches include hospital bedside reactors and cell "platforms" that can be mass produced and then tailored as a final step.95 Significant efforts in developing efficient manufacturing processes and supply chains for personal genomics will continue.
Likewise, in AI, one of the biggest challenges is developing unsupervised learning. Today, the engineer is heavily involved for much of the algorithm development on a given data set. Scalable, deployable AI will be enabled by algorithm development that can happen without direct involvement of the engineer, or “unsupervised.”96
As companies prepare for the floodgates to come down (particularly for AI), they will race to install test systems and build infrastructure. We’re likely to see an inundation of niche AI technologies focusing on individual functional areas and heavy capital investment to deploy algorithms. Most experts expect faster deployment of AI in manufacturing, supply chain and consumer-facing marketing.97 In parallel, expect an increased awareness of the value of data and, therefore, increased competition in data access as companies jostle for advantage. As large companies race to get a foothold in suddenly exploding industries, expect more large conglomerates to acquire start-ups, particularly in personalized medicine.
In four to five years, the floodgates will really begin to come down, particularly for AI as a result of large-scale deployment of infrastructure. Companies that have spent the time to develop talent and infrastructure will begin to see accelerating returns – and companies that have lagged will struggle to catch up. In parallel, the personalized medicine and battery development markets will start to solidify as large companies strategically scoop up start-ups and emerge as dominant forces.98