Knowledge

Artificial intelligence and machine learning — Debunking the myths and defining today’s reality

Artificial intelligence and machine learning — Debunking the myths and defining today’s reality main image

At the largest congregation yet of the Whitespace Innovation Community, the focus was on artificial intelligence and machine learning, as well as attempting to separate the hype around them from reality and potential.

Meeting Theme

At the largest congregation yet of the Corporate Innovation Club, the focus was on artificial intelligence and machine learning, as well as attempting to separate the hype around them from reality and potential.

Often confused, amalgamated and misunderstood, the two related technologies have recently enjoyed much attention via their newfound status as buzzwords, joining the ranks of VR, drones, and blockchain.

Advances in AI and ML are presently burning bright with potential.

Existing as the subject of a buzzword serves as a double-edged sword, of course. AI and ML – as they are more commonly known – are enjoying much attention today. That both spotlights and accelerates their development and potential. There is no smoke without fire – or so the proverb insists – and advances in AI and ML are presently burning bright with potential. But with hype comes a disorderly rush to board the bandwagon, misrepresentation via overenthusiastic marketing speak, and even cynicism from jaded observers all too often burned by embracing emerging technologies with good intentions.

Fitting and important fodder, then, for a spirited and open-minded Corporate Innovation Club get-together.

Key Takeaways

What differentiates AI and ML?

AI and ML are understandably frequently confused. They are both, after all, technological concepts connected to the broad notion of computers possessing the capacity to make independent decisions and ‘think’ for themselves. AI is the umbrella term here, and refers generally to computers’ – and before them other mechanical devices’ – ability to process tasks using their own intelligence and decision-making capacity. Importantly, the definition of AI continually evolves. As we understand more about both human intellect and computing power, the potential of what AI can be expands and shifts.

Machine learning, meanwhile, is a specific application of artificially intelligent technology to handle and learn from large bodies of data. Machine learning may let a computer choose its own most efficient way to handle a specific task, or let it identify patterns and trends in big data sets that can improve consumer experiences.

Today we are in the opening stage of a three-part evolution of what AI can do

Experts in AI see that the ultimate evolution of AI – carrying ML with it – will take a three-stage form, at least as far forward as is reasonably predictable.

  1. First comes ‘narrow AI’; the position in which we find ourselves today, where AI can be precisely applied to specific functions.
  2. We are approaching an era of ‘general AI’ that will see computer thinking take on a comparable form to human thinking. Such AI will be flexible, dynamic, subtle, and able to make nuanced decisions, with less input and guidance from human agents if necessary.
  3. Finally, we will arrive at ‘super AI’; something we are currently nowhere near, where computer intelligence will vastly outperform human intelligence. The outcomes of such AI on society, culture, work, and environment are near impossible to precisely predict, but the arrival of super AI may disrupt like no technology before it.

Don’t fear advances in AI and ML

We have a long way to go before we reach a point where science fiction’s vision of AI is anything like possible. However, there was some cynicism in the room about the power of AI to disrupt not just industries, but humans’ roles and work in those industries. It was noted, though, that we should consider the current and next generations of AI as a technology to augment human process; not replace it. AI remains far from perfect, and far from capable of understanding human nuance, or non-logical thought processes. It processes information without instinct, gut feeling, and compassion. The example was given that in the HR and recruitment space, AI would struggle to pick a perfect graduate; but the technology may be used to monitor and counter the inherent bias that exists in every human. Such a pairing of human intelligence and artificial intelligence is now commonly referred to as an ‘augmented intelligence’ technology.

We should consider the current and next generations of AI as a technology to augment human process; not replace it.

Ethics committees are already well established as a means to debate and guide the use of emerging technologies like AI and ML, particularly in academic contexts. More questions must be asked, however, about where in a Corporate structure an ethics committee can best serve its purpose.

For Further Consideration

  • Will we benefit from AI and ML with more potential if we can first install more diversity in the Corporations adopting it, and the panels, committees, and teams shaping it?
  • What responsibly do investors in AI and ML have in guiding ethics, or working with Corporates to do so?
  • Should we be worried that some AI experts are allegedly noticing that bots are unfollowing them on social media? This example was presented playfully in one of the meeting’s lighter moments but presents an interesting thought experiment to explore oneself.