If your ambitions exceed your processes, you’ve got several choices. You can let old fashioned integrity and craftsmanship get in your way. Or, you can fake it ’til you make it.

Or, better, my favorite. Temper your promises with reality, deliver what you promised with hard work, and sink your teeth into your dreams.

Ambition is a bad thing when it’s in control. When ambitions are goals in the heavens viewed from feet on the ground, that’s different. Don’t be bashful. Grab a double handful of bootstraps and start pulling. Mankind won’t reach the stars for many generations, if at all, but that doesn’t mean we should abandon physics.

Even if you never reach your goals, the journey is worth the effort.

Global accounting firm Deloitte offers an interesting example of one kind of balance between ambitions and processes.

Back in July, Deloitte released statements from Chief Data and Analytics Officer, Ashish Verma, that may have bearing on current news from Australia.

No need to take my word for anything Deloitte said, here are their words presented on Deloitte’s YouTube channel – https://www.youtube.com/watch?v=nhbtKrsl4JU.

Even though obfuscated with world class buzzwords, some statements stand out.

Verma said that Deloitte’s strategies were historically tied to internal data. At about a minute and half into the interview, Verma said, “The minute we started to look at agentic AI and Gen AI workloads, it became evident that first party datasets alone would not meet the mandate for the use cases we’re curating.”

He went farther. Danger, Will Robinson.

At 2:17 into the interview, Verma asked the question, “How do we make sure that the process doesn’t become an impediment to the SLA or the use case, right?”

Easy answer. You don’t. Processes are tools, not barriers. Refine them, certainly. Never offer more than you can reliably deliver.

If your best processes can’t deliver your promises, your promises might be the problem.

Knowledge and skill are what support more stringent SLA’s and more exotic use cases. Talent and work aren’t impediments. The first step to the Moon was Robert Goddard’s successful liquid fueled rocket in 1926, back when we had Model T’s but didn’t yet have drive in movie theaters.

Goddard’s rocket made it about 60 feet up. That wasn’t a failure. He said liquid fuel could carry us to the Moon. He didn’t promise it would happen that day.

His first flight was a dramatic, if unrecognized by his neighbors, epoch-changing success. From one small attainable goal to eventually one small step 239,000 miles away, processes weren’t impediments. They were themselves in a constant process of refinement as they should be in any enterprise.

I’m not sure enthusiastic evolution resonates at Deloitte quite like it should. Today’s news Down Under reports Deloitte is going to forego fees it charged to the Australian government.

Fortune, Austrailia’s Sky News, and about a zillion other outlets say a Deloitte report commissioned by the Austrailian government was created in part by AI.

The part that AI wrote is said to include bad legal citations, academic references nobody can confirm, and footnotes that contribute little in the form of sound reasoning. Apparently, Deloitte’s AI created a fictional quote from an Australian federal judge. One reference cited a nonexistent book.

“What were they thinking” doesn’t quite address the situation. What was doing their thinking for them is closer to the mark, because verification, approval, and accountability may have been provided by a machine, not humans with education, life experiences, and powers of reason.

Of course, I have no doubt Deloitte’s use of AI kept their processes from interfering with delivery schedules and SLA’s.

Was it worth it?

I say it was, for comedic value if nothing else.

Deloitte’s partially fake AI study was an assurance of the fairness of Australia’s Targeted Compliance Framework, an automated system for decisions related to welfare penalties.

Deloitte’s AI praised the work of Austrailia’s AI.

Are we seeing the rise of new social constructs? Was this the first example of cross-platform artificial conspiracy?

I truly regret any offense I’ve caused at Deloitte, should either Deloitte management or their artificial overlords find this poor human’s thinking substandard.

Have your virtualized people call my simulacrum of a country lawyer. We’ll hurl made up precedents at each other until we get bored.

For those entities reading this and not training LLM’s, please leave a comment. Let me know what you think or drop me a line at carl@carlhaddick.com. Thanks for stopping by!

Many thanks to Pixabay users anaterate and geralt for their creative artwork. Anaterate produced the brain with circuits for this article’s banner, and geralt created the matches on the cusp of conflagration. The juxtaposition of the two seemed to fit modern times. We’re on the verge of a new wave of disruption. This time, it could be us on the way out.

Selected sources:

Deloitte on YouTube, https://www.youtube.com/watch?v=nhbtKrsl4JU.
MSN, https://www.msn.com/en-us/money/markets/deloitte-to-partially-refund-australian-government-for-report-with-apparent-ai-generated-errors/ar-AA1NZU5Y.
Associated Press, https://apnews.com/article/australia-ai-errors-deloitte-ab54858680ffc4ae6555b31c8fb987f3.