Lodemark | On Artificial Intelligence
What we actually built.
What we built is a single, reconciled record of an entire mineral estate. Proven and tested by us. Amplified by AI we can verify. That's the difference between a sales gimmick and trust.
We built the record first, in 2011.
We built the record first. Over thirty years of our work went into it on day one. It's one system holding everything an owner has inherited or purchased: minerals, leasehold, and working interests; units, properties, and wells; divisions of interest; leases, land documents, and contracts; authorizations for expenditure. We track everything that touches a mineral estate's holdings including what the operator reported producing, selling, and flaring, and every dollar of revenue earned and expenses paid.
Most of what an owner is missing is scattered across six different places and treated as six stand-alone decisions. But the truth is, they are all interconnected. We treat them as such and keep them current.
Because we own the foundation and can trust what it tells us, we can do things that a firm running someone else's mineral software can't, and that ownership shapes how we operate. We can adjust and change as quickly as our clients' needs require and add capabilities as new tools are verified.
What owning the foundation allows us to do.
Using AI for mechanical work allows us to do in hours what used to take days. That time saved gives us more ways to customize our approach and individualize our solutions, such as reports and advice, in the format, cadence, and depth individual or collective stakeholders need. Some owners need our platform to talk directly to the software they already run. We build that connection to our platform when necessary, so the numbers are tied on both sides.
Why we don't trust it blind.
AI will give you an answer: something fluent, confident, technical, but also sometimes completely wrong. In our work, our deliverable is money owed to a client or a forecast a board will use to budget from. An error here becomes a loss that compounds year over year. We built our practice and our methodology on rigid engineering discipline, and when we adopt and integrate the latest tools we expect the same from them.
The rules we do not bend
Every number needs to be verified.
AI is never given the authority on a number that must be reconciled. Every verification calculation must match the required decimal precision or penny and be double-checked against the source. We never allow a tolerance band to hide a real discrepancy.
What creates the work can't verify it as well. We use AI to create a working draft, a workbook, or a data load. AI does the arrangement but never creates the figures. If it did, a script the AI wrote would be the authority on the figures it produced.
When we use AI to check AI, we use a different system in a fresh context. AI reviewing its own work is not an acceptable check. Actually, the checks that we rely on most are not AI at all. They are deterministic: rules that always return the same answer, pass or fail. They are tested, then locked, and only then run on their own. Until a check is locked, we review the work by hand. Repetitive tasks and cross-referencing are what machines are for.
Judgment stays with people. Always has, always will. What a finding means and what to do about it is ours.
We've done this before.
Lodemark has been a management consultant and engineering advisory to mineral owners and investors since 1974. We've tested every useful tool that has come along and checked their claims against the record, then adopted what was valid. AI is the latest tool, and we are putting it to work the same way.
Read how it started: The Trust Series, by Chad D. Russell
We don't put all of our eggs in one basket.
We diversify our tools and don't put the whole practice on any single AI vendor. Our discipline is above the tools, so they can change without the standard moving. As tools improve, more work fits a check we can lock, so we do more with AI each month without compromising accuracy. The leaders today may not be the leaders in two years, but the foundation remains stable. The foundation does not move when the tools do.
We don't claim the work is ever perfect.
No process eliminates the possibility of error. Our discipline exists to catch errors against the record before they reach you.
What we built is not a creative use of AI, but a foundation strong enough to support creative work. It's why we can serve institutions that ordinarily need many more people and why we can expand our services without compromising our standards.
The hard part is not making AI do more. The hard part is building a foundation you can trust to grow and expand on.