Advertisement

2023 Predictions: Breakthroughs Incoming for AI and Data Science

By on
Read more about author Phil Tee.

Surprising though it may sound, most AI business applications remain rudimentary. Some industry spectators believe the market is flush with cutting-edge AI breakthroughs. But in truth, many providers mislabel rules-based tools – including robotic process automation (RPA) – as AI and ML. 

Seasoned DevOps teams and software engineers know that RPA – although useful in its own right – is not AI. RPA follows pre-established rules instead of adapting and learning with its environment. Adaptability remains a classic tenant of true AI. Without it, systems cannot grow and intuit new solutions – they remain woefully stagnant alongside existing data.

Nevertheless, AI and data science sit at a fascinating juncture. Practitioners who live, breathe and sleep AI/ML have made incredible strides over the past five years. We see that progress coming to fruition over the next year, opening new doors for software engineers, medical professionals, and quantum physicists alike. 

Let’s unpack how AI has improved and discuss how its transformation will soon yield life-changing results.

AIOps will enter a new era of vertical integration

Like RPA, many AIOps offerings on the market masquerade as intelligent and self-learning solutions. However, most products claiming AIOps technology focus only on events and/or application performance monitoring – just a small part of AIOps. These tools prioritize one dimension of AIOps over all others, creating an incomplete and problematic view of the incident lifecycle.

It is my firm belief that this approach will not stand the test of time. Only AIOps tools encompassing all four classic telemetry inputs – from detection to resolution – will achieve the holy grail of self-healing systems. True AIOps tools radically reduce mean time to recover (MTTR)/mean time to detect (MTTD), which creates less system downtime and considerable cost savings.

Speaking of cost savings, IT leaders are on the hunt for processes that simplify bloated IT service management (ITSM). We increasingly see AIOps as a solution here. When a system detects a data anomaly, AIOps can create an automatic ticket – with all relevant context – for human administrators to decipher. Additionally, the system can learn via ML, constantly improving its understanding of human-generated tickets and escalating ITSM processes appropriately. This functionality reduces repetitive tasks that steal time from high-priority infrastructure initiatives. Crucially, with better fault isolation, the number of tickets that require processing can be reduced dramatically – often by more than 50%. Moreover, a new collaborative approach to problem resolution can squash an organization’s MTTX.

Innovative AI applications in medicine will save lives

The doubling time for AI sophistication – so to speak – has decreased significantly in recent years, accelerated, no doubt, by the pandemic and overall digital transformation. We are gaining momentum at a rapid clip.

Arguably the greatest promise for rapid AI upscaling lies in the medical field. The next few years will see monumental progress in medical treatments. In fact, we are already seeing early signs of a breakthrough: Dr. David Baker, the 2021 Breakthrough Prize winner, used AI to design completely novel proteins. This ground-breaking technology has several possible applications in the life sciences and may ultimately lead to the creation of life-saving medical treatments for diseases like Alzheimer’s and Parkinson’s.

Quantum foundations may create a third wave in AI

Although the challenges involved in building a practical quantum computer are phenomenal, the crossover from fundamental physics into informatics has created a promising level of quantum-inspired computing. QuantrolOx, which uses AI to tune quantum computers, offers a compelling example of possible future applications. This crossover will also facilitate new techniques in data science. We’re especially interested in the exploration of areas of math other than discrete linear algebra, such as algebraic topology and quantum foundations, which could become a new frontier in AI technology. This innovation may even herald a third wave in AI.