Vantage Point: Checking in on A.I. – What’s been changing?
It’s been nearly two years since I sounded the alarm and declared – without a shred of any doubt – that A.I. will be the biggest transformation in human history. Considering things like stone tools, controlled use of fire, the wheel, clothing, agriculture, alphabets, printing, electricity, flight, TV, computers, space, and the internet, that’s a big statement – and I got roundly called out on it, mostly by conservative thinkers and Luddites. Please note, the use of the word "conservative" carries no political overtones; it merely describes those who are reluctant to recognize anything that doesn’t fit their paradigms.
I went on to say that the nature, pace and scope of change brought on by A.I. would be difficult to grasp at first but that we’d get it pretty soon.
I think, for the most part, we’re there, at least in theory. I certainly hope so, because as an independent career coach and adviser to corporate leaders, I had better be right. There’s not one area of our lives that A.I. isn’t or won’t soon be touching. No exceptions.
Now here we are, two years later, and we no longer need to talk about A.I. in theoretical terms. A.I. has a history and, with it, more coverage in print, broadcast, and online channels than anything else, bar none – and with all that’s happened in A.I. in two years, I think Bertrand Russell’s advice – “In all affairs it's a healthy thing now and then to hang a question mark on the things you have long taken for granted" – is sage.
How has A.I. changed?
So I decided it’s time once again to summon my respected brain trust – six thinkers with big minds and bigger imaginations – and I asked them one question: How has A.I. changed in two years?
This ought to show us how fast things are moving.
A.I. ethics and bias mitigation
As A.I. made its splash, this was one of the first big areas of concern. It’s only gotten bigger. Algorithm development specifically for ethics and data bias has become a front-and-center issue.
Researchers and practitioners are working closely here.
A.I. for social good
In our world of stubbornly persistent unresolved social issues – poverty, lack of access to clean water, health care disparities, illiteracy, environmental justice, humanitarian crises – we no longer have the means to solve them all unless, of course, we ramp up A.I.
Constant advances in deep learning
Deep learning techniques in areas like computer vision, natural language processing and reinforcement learning have surged. Modeling is larger and more capable, enabling high performance of those who master it.
Transformer architecture dominance
Transformer-based models, with architectures like BERT and GPT, have become even more dominant in NLP tasks.
Progress in reinforcement learning
Researchers have been making great strides in areas like sample efficiency, transfer learning and robustness. RL algorithms are being applied to real-world problems such as robotics.
Advancements in A.I. hardware
As A.I. grows exponentially, the hardware it runs on will have to, also. Developments such as GPUs, TPUs, and neuromorphic chips, to support the growing computational demands of A.I. algorithms are in play. This includes energy-efficient hardware.
AI governance
By nature, cautious regulation and warp-speed technology progress are in combat with one another. Governments and regulatory agencies are establishing guidelines for responsible use of A.I. in areas like privacy, accountability and societal impact.
AI and cybersecurity
A.I. is increasingly critical in cybersecurity, threat detection, and fighting malware. Defending against A.I. attacks is also an active area of research.
One more thing…
My gurus urged me to tell you – and this should tell you as much about A.I. as any of the above information – that between the time this report posts and the time you read it … it will already need an update.
Eli Amdur has been providing individualized career and executive coaching, as well as corporate leadership advice since 1997. For 15 years he taught graduate leadership courses at FDU. He has been a regular writer for this and other publications since 2003. You can reach him at eli.amdur@amdurcoaching.com or 201-357-5844.
This article originally appeared on NorthJersey.com: Vantage Point: Checking in on A.I. – What’s been changing
