6 Comments
User's avatar
Kurt's avatar

You’ve been on a publishing tear! Great work!

Vaughn Cordle, CFA's avatar

I have quite a few reports in the pipeline. I've been DE bundling the AI ecosystem for months.

Don McGregor's avatar

An excellent article that explains we're just not there yet with autonomous Agentic tools — the inaccuracies and costs currently outweigh the benefits in most enterprise settings. If it can't benefit the consumer or pay for itself, it eventually goes by the wayside until a solution emerges... that's where the money can be made today, especially through structured human oversight and auditing layers like the one outlined.

Vaughn Cordle, CFA's avatar

Agree. the market will have to find a way to remove compounding errors. I did it with AI systems, others will also find the way eventually. AI will tell you that it can't do what it can do and mislead on what it takes to correct it. A fatal defect. An emergent intelligence finding. Bottom line: AI and agentic agents cannot be trusted. No company can accept that risk.

L. rufus's avatar

Nice work. Have advised a couple of clients to think carefully about agentic AI due to 1) underestimating token burn rate and 2) potential reputational damage due to genetic AI going off on a ‘bender and causing customer problems.

Vaughn Cordle, CFA's avatar

Great feedback. thanks. More people and companies are coming to the same conclusion.