logo
Published on

The GenAI Terminus

Authors
  • avatar
    Name
    Strategic Machines
    Twitter
stack

Working at the Edge

If you have been tracking with us on the remarkable journey in GenAI, you'll have noted a pretty consistent theme. We find the technology astonishing, up to a point. It fits well for some use cases, and others, not so much. Our most recent posts, based on our work with clients, tried to point out the limits, while highlighting the opportunities.

Dr. David Ferrucci, the CEO of Elemental Cognition is a visionary in the field who understands not only the limits of GenAI, but where those limits reside. He recently noted:

... for the same reason [LLMs] are creative expression generators, [they] are not the right tool to deterministically and reliably handle the complexities that proven, mathematically precise algorithms can. ... While statistical, generative processes like LLMs can be wondrous, alone they are not well-suited for driving repeatable, deterministic processes, especially for regulated industries where transparency and reliability are essential.

We would only add that we think LLMs are not appropriate for "driving repeatable, deterministic processes" in any Industry.

Dr. Ferrucci has important guidance for any team weighing potential use cases for GenAI. It is always valuable to push to the edge for 'needle moving' impact with technology, just as long as you know where the limits reside.