REAL Talk With Sam Holcman

We have compiled the six most often asked questions when we are called on for a potential consulting engagement. These six questions address the most important skills a consultant needs.  And you should be able to gauge the answers to these questions quite quickly.  And yes, there is one time-waster question we also suggest really is a time waster!

In a recent article in the Wall Street Journal, Paul Vieira outlines the actions taken by the Canadian Government that would force Facebook and Google (amongst others) to pay media outlets for links.  I call this leveling the playing field, and this podcast outlines why it is time for this to happen.  And by extension, the same compensation needs to occur from the AI vendors that are training their tools on intellectual capital they do not have rights to.

An article I read - "7 Ley questions CIOs need to answer before committing to generative AI," got me thinking about what questions need to be ASKED prior to not just AI - but any technology that is acquired/implemented. Using just a small component of the EACOE Enterprise Architecture approach will provide these questions. It is worth just a few minutes of your time, I believe.

Whether you are into AI or not, we all have a choice to make when we are looking for information of any kind – restaurant recommendations, business trends, gasoline prices, strategies, or best practices.  Two fundamental sources exist – paid sources and no-cost sources.  Have you thought about it in these simple terms?  That is what this broadcast is about.

I recently read an article titled AI is Writing Code Now. For Companies, That is Good and Bad. In summary, it is bad.  It is not only bad because we are increasing complexity and exacerbating technical debt, but these organizations are just speeding up what they have been doing for decades. It is time to realize where we are in Information Technology Maturity and get the organization on the path to Maturity Level 3 – Assemble Solutions to Order.

In a recent CIO article, eight famous analytics and AI disasters are described, discussed, and analyzed.  There is a common theme that is apparent after a bit of analysis is done.  I call this the ninth and continuing disaster.  A solution to this ongoing disaster is outlined.

As a public service, I am reading the actual “Bard Privacy Notice” in this podcast.  It is worth all to spend five minutes listening to this Notice.  I will then relate this Notice to what it means to those of us that in some way, rely on the Internet/AI for information and the use of this information.  WOW!

On the heels of a widely published academic study on the AI-enabled cheating occurring on college campuses surrounding exams, college campuses are studying and looking at alternatives – both oral exams and, more importantly, skills-based demonstration, to access academic achievement and competency.  The motivation is quite simple for colleges.  If a degree no longer designates legitimate achievement, it will cease to hold value.  Colleges are at a real risk of becoming diploma mills.  It is time the Enterprise Architecture and Business Architecture certification community recognized this situation before certifications in these fields became what some are already calling "certification mills."  Some certification bodies are already there.

The fuel for Artificial Intelligence is data.  Good fuel (data) will provide good energy (answers).  However, bad fuel (bad data), will cause issues both in the long term and short term (bad answers, misleading information, untruths, and many more undesirable outcomes).

There is no known, as of this broadcast date, list of data sources that any of the AI engines are being trained on.  What does that mean to the AI “consumer”?  Maybe we need a different approach to how to train these AI engines.