During a recent conversation with a client about how fast AI is advancing, we were all struck by a point that came up. Namely, that today’s pace of change with AI is so fast that it is reversing the typical flow of innovation from a chase mode to a catch-up mode. Let’s dive into what this means and why it has big implications for the business world.
The “Chase” Innovation Mode
In the realm of analytics and data science (as well as technology in general) innovation and progress have historically been constant. Furthermore, new innovations are typically seen on the horizon and planned for. For example, it took a while for GPUs to begin to realize their full potential for helping with AI processing. But we saw the potential for GPUs years ago and planned ahead for how we could innovate once the GPUs were ready. Similarly, we can now see that quantum computing will have a lot of exciting applications. However, we are waiting for quantum technologies to advance far enough to enable the applications that we foresee.
The prior examples are what I mean by “chase” innovation mode. While change is rapid, we can see what’s coming and plan for it. The innovations are chasing our ideas and plans. Once those new GPUs or quantum computers are available, we’re standing by to execute. In a corporate environment, this manifests itself by enabling an organization to plan in advance for future capabilities. We have lead time to acquire budgets, socialize the proposed ideas, and the like.
The “Catch-up” Innovation Mode
The advancements with AI, and particularly generative AI, in the past few years have had a breathtaking and unprecedented pace. It seems that every month there are new major announcements and developments. Entire paradigms become defunct practically overnight. One example can be seen in robotics. Techniques were focused for years on training models to enable a robot to perform very specific activities. Enabling each new set of skills for a robot required a focused effort. Suddenly today, robots are using the latest AI techniques to teach themselves how to do new things, on the fly, with minimal human direction, and reasonable training times.
With things moving so fast, I believe we are, perhaps for the first time in history, working in a “catch-up” innovation mode. What I mean by that is that the advances in AI are coming so fast that we can’t fully anticipate them and plan for them. Instead, we see the latest advances and then must direct our thinking towards understanding the new capabilities and how to make use of them. New possibilities we have not even thought of become realities before we see it coming. Our ideas and plans are playing catch-up with today’s AI innovations.
The Implications
The pace of change and innovation we’re experiencing with AI today is going to continue and there are, of course, benefits and risks associated with this reality.
Benefits of catch-up innovation
- Nobody can see all that will soon be possible and so organizations of all types and sizes are starting on a largely equal footing
- The availability of new AI capabilities is broad and relatively affordable. Even smaller organizations can explore the possibilities with today’s cloud based, pay as you go models
- In some cases, smaller organizations can bypass traditional approaches and go straight to AI-led approaches. This is similar to how some developing countries bypassed implementing (and transitioning from!) traditional landline infrastructure and went straight to cellular phone service
- Organizations win by continually assessing needs versus capabilities because what wasn’t affordable, or even possible, a short time ago may now be easily accomplished for cheap
Risks of catch-up innovation
- The deep pockets of big companies won’t provide as much an advantage as in the past and large companies’ organizational momentum and resistance to change will provide opportunities for smaller, nimble organizations to successfully compete
- With AI’s self-learning capabilities rapidly advancing, the risk of harmful or dangerous developments occurring increases greatly. We might not realize that a new AI model can inflict some type of harm until we see that harm occur
- Keeping current is even more overwhelming than ever. Major technology, AI, and analytical process investments may be outdated even before they are completed and deployed
- On both a personal and corporate level, the risks of falling behind are greater than ever while the penalties for falling behind may be higher than ever as well
Conclusions
Regardless of how you interpret the rapid evolution and innovation in the AI space today, it is something to be acknowledged. It is also necessary to put concerted effort into staying as current as possible and to accept that some strategies and decisions made given today’s state of the art AI will be outdated in short order by next month’s or quarter’s state of the art AI.
Since we’re in a novel “catch-up” innovation mode for now, we should try our best to take advantage of the new, unexpected, and unplanned capabilities that emerge. While we may not be able to anticipate all of the emerging capabilities, we can do our best to identify and make use of them as soon as they emerge!
The post Has AI Changed The Flow Of Innovation? appeared first on Datafloq.
