10. Automated warfare may begin in 2019.
Modern automated warfare began with the Aegis Ballistic Missile Defense System, which has roots in the 1980s Reagan era. The potential now exists for high-scale automated offensive attacks, whether used by terrorist groups or state actors at massive scale. The risk of a billion drones attacking complete with primitive coordination is real. A prudent New Year’s resolution for the DoD would be to prevent an automated version of Pearl Harbor.
9. Automated cyber attacks.
This trend is sadly already well underway. The sophistication and scale are increasing at an exponential rate while substantially outpacing automated defensive measures. Leaders in the private and public sector may finally realize in 2019 that no amount of training will suffice. Throwing more troops at a Gatling gun didn’t work terribly well in the 1800s. Humans are simply no match today for automated cyber attacks. Next generation AI systems must be adopted in 2019 or we are likely to see sharply increased churn in senior management and boards, if not corporations and nations.
8. First responders transition to preventers.
Price and performance improvements in AI components combined with improved algorithms are rapidly expanding prevention capabilities across society. Over the course of the next 12 months we should see a healthy increase in prevention of fatalities from semi-autonomous systems. A few examples include drone delivered resuscitation, health device monitoring, wildfire mitigation, industrial accidents, and semi-autonomous vehicles that reduce traffic accidents.
7. Smart infrastructure begins to scale.
The U.S. is falling behind in smart infrastructure due to long-term underinvestment by the federal government. States, counties, and cities are partially bridging the gap by increasing local investment. However, most projects are not yet designed for intelligence and prevention capabilities, meaning it will cost even more to retrofit at a later date. National infrastructure is one of few areas where bipartisan support exists, so hopefully we’ll see a rational smart infrastructure bill approved in 2019.
6. Higher education begins a long journey of transformation.
In 2018 we observed the expansion of the student debt crisis and calls for a federal government bailout. The $22 trillion question (and growing rapidly) is who will bail out the government? Serious competition and new models are just beginning to emerge from within existing universities and new pathways. Completely new models may eventually displace higher education unless the system becomes competitive and accountable. Look for an increase in AI-enhanced life-long learning programs that provide higher quality at greater scale and much lower cost.
5. Healthcare rediscovers the Hippocratic Oath with the help of augmentation.
That the U.S. healthcare system is fundamentally broken is undeniable. Our costs are twice as high as other countries that enjoy higher quality healthcare (U.S. isn’t even in the top 10). The complexity in the healthcare system as well as individual health is beyond the ability of humans alone. The amount of information required is simply too vast to manage well unless machine intelligence is applied with precision. The opioid crisis is just the latest in a long history of readily preventable large-scale catastrophes in healthcare. Systems have existed for many years that could have easily alerted state and federal authorities to this terrible disaster. Millions of highly addictive pain pills distributed to small town pharmacies is a simple anomaly to capture even with a dumb system. If healthcare organizations don’t get serious about adoption of advanced AI systems they will be displaced. The U.S. will be insolvent within decades otherwise.
4. Ratings of public and private sector entities will be influenced by adoption of AI systems.
This has been a personal priority for the last three years in pioneering our HumCat program and product line. By adopting state-of-the-art AI systems, government and corporate entities can not only improve their risk profiles, but the reduction of risk can be demonstrated. The result should be improved ratings and lower borrowing costs, which is becoming very important given normalized interest rates and bond bubbles. In contrast, those organizations that don’t adopt state-of-the-art AI systems can expect to see lower ratings and higher lending costs in the near future. They may also experience a brain drain as smart people and organizations generally prefer to live and work in areas with a sustainable trajectory.
3. Boards suddenly realize the vast scale of redundancy and waste in their AI experiments.
Leading consulting firms have generally done a good job on AI systems reporting in the last few years. We’ve noticed a significant improvement in quality and have attempted to help. However, one area that still consistently fails clients is advice to start slow and learn in incremental fashion. Very few organizations enjoy the luxury of time or can absorb the cost of redundancy as organizations compete in the talent war just to perform almost identical AI tasks. In addition, AI infrastructure in cloud leaders provides no competitive advantage whatsoever as their systems are available to everyone globally. Meanwhile, productivity, performance and competitive gaps between leaders and everyone else are expanding rapidly. This translates to the necessity to embrace emergent leaders that provide a competitive advantage. No different than previous generations.
2. Traditional incumbents form alliances with emerging AI leaders to level the playing field with big tech.
During the first phase of the recent machine learning revolution, which started about six years ago, the super majority of investments in new companies were for strategic control or acquisitions. Governments and corporations alike deemed AI a must-have survival tool. Unfortunately, this also resulted in China rapidly pulling ahead of the U.S. and EU in AI adoption as incumbents refused to support the next generation of critical domestic technology. China correctly viewed this behavior as an historic opportunity to leap frog the West. In 2018, even AI leaders like Google are finding it necessary to form alliances and become customers rather than attempt to own or control. Major corporations and governments are now investing a significant portion of the total in AI as customers and partners. We expect this trend to continue to expand in 2019.
1. Job displacement will become secondary to organizational, sector, national, and species displacement.
While fear of displacement is justified for a few specific jobs over the next five years, such as truck drivers, most of the hyped reporting on jobs displacement has been blatantly biased. A retraining policy for specifically impacted jobs such as truck drivers is I think prudent and humane, perhaps patterned after Germany’s unemployment program that splits the costs between government and employers. However, over the broader economy increased efficiency has resulted in more jobs, not fewer, and at higher compensation.
That said, significant risks do exist in the AI revolution. The greatest risk from AI facing companies and nations is slow adoption of advanced AI systems and/or poor execution. The fast follow approach that became such a common strategy over the past two decades with incumbents is failing fast in AI systems. An incremental approach is more than just high risk behavior during revolutionary change–it’s certainty of failure. Indeed, the “let’s wait and see” strategy in the case of AI systems should be viewed as suicidal, whether driven by a corporate Luddite culture, incompetence, or complacency.
May you and your organizations make wise decisions in the coming year. Wishing everyone a merry Christmas, happy (and healthy) New Years, and the best possible outcome for your organizations in 2019 – Mark Montgomery.