CTO View: 11 Predictions for 2030
There have been plenty of 2020 Predictions lists out recently, and they have all been interesting to read. As a tech company CTO, I have a special responsibility to look further ahead, so here is what I am predicting we will see, not over the next 10 months, but over the next 10 years. With input from my team of technologists, I present to you my predictions for 2030.
1. ‘Approximate Computing’ goes mainstream. Approximate computing will become the standard approach to the large-scale simulation/analysis of natural and some virtual systems. This involves turbo-charging approximate analysis with cloud-delivered access to Quantum-based accelerators that function more like old analog computers. The application of quantum computers to more precise analysis continues to be stymied by physics (i.e., hardware error rates), the long intervals between algorithmic breakthroughs, and continuous improvement in the (competing) heuristics used with classical computers.
2. Popular film stars are virtual. By 2030, over 50% of major motion pictures will feature virtual stars/co-stars. These completely life-like characters will have unique personas and individual voice signatures just like their human counterparts, but they will be wholly digital creations with an entire digital identity and millions of fans.
3. Targeted Intelligent Medical Diagnosis. Many medical diagnoses will be greatly enhanced due to machine learning techniques using clinical data to make highly accurate medical diagnosis and recommend custom-tailored treatments specific to the genetic and medical history of patients.
4. Classical Computing gets a temporary reprieve. After years of flatlining performance, classical CPUs get a boost in the late 2020s as engineers realize that considerable performance had been “left on the table” when Moore’s Law was speculated to be slowing. Since these improvements are somewhat ad-hoc (mostly based on new materials, transistor structures and cooling), they do not restore the steady drumbeat that Moore’s Law provided but promise to yield a 4-8x improvement in the early 2030s.
5. The post-DNN era dawns. By the mid-2020s when the limitations of Deep Neural Networks become evident, DNNs will be replaced by superior new techniques whose underlying theory is well-understood, and which require vastly reduced training resources. Nonetheless, the DNN momentum continues, driven by the billions of dollars already invested.
6. 5G and its successors blanket the planet. By 2030, 5G and its successors will reach into every corner of the world, bringing access and information to everyone and everything. New social structures will emerge due to global connectivity anytime, anywhere, anyplace.
7. Adoption accelerates for self-driving vehicles. By 2030, 20% of highway distances in the US/Europe/China will be driven by autonomous vehicles, triggered by a ~2025 revolution in the structure of long-distance trucking and local delivery businesses. The transition is even faster for agricultural vehicles, like tractors and harvesting equipment. Consumer adoption gets a late-in-the-decade boost from new state incentives to drive adoption by octogenarians.
8. ‘Serious Developer’ culture becomes a thing. Over the next decade, ‘Serious Developer’ culture gets widely embraced by professional software developers in an ethics-driven response to the late 2010s when “nowhere near beta” code was routinely shipped into systems affecting billions of consumers. As a nurturer of a values-driven culture, VMware will be among the first to aggressively empower serious developers with its Tanzu tool suite.
9. Energy storage gets a boost. Improvements in battery charge densities – more stored energy at lower cost – continue to evolve through the 2020s. Although battery improvement has been slower relative to other technologies, the compounding effect of technological improvements enables dramatic gains over carbon-based energy systems. Energy storage is no longer the barrier to widespread adoption of renewables as a preferred energy source, resulting in a profound impact on the design of many types of energy-dependent systems.
10. Climate change action: better late than never. With multi-lateral systems unable to effectively address global climate change, others will step up in response to climate-driven natural disasters. As with privacy, the EU plays a leadership role; the US effort is led by large enterprises partnered with local/state governments, while China sees a grassroots “harmony with the environment” movement. Emissions finally begin to level off due to reductions in fossil fuel dependence, aided by improved energy storage and carbon-capture technologies.
11. Distributed Ledgers go mainstream. Distributed ledger technology continues to evolve, becoming a dominant form of secure information-sharing across organizational boundaries. Digital contracts for many forms of multi-party agreements (e.g., financial securities) will facilitate more efficient global economic commerce.