Inspired by a combination of the Miami Sun and being
surrounded by brilliant technical minds this past week, I am compelled to
expound here on a topic of concern. About a year ago I opined that
technology’s rapid development was outpacing the ability to regulate its
use. I proposed a mathematical formula to help track the phenomenon.
Despite the brilliance and extreme talent of some of the people I am blessed to
know in the technology sector, I am concerned this gap is widening at a
frightening pace.
Since it has been a year since my original post setting
forth a formula to determine if my concerns are justified, I am providing a
one-year comparative update. Providing a brief summary of my original proposed
formula, here is a recap of the concept:
For simplicity, the original formula considered these
following components:
1. Pace of Innovation (I): This is a measure of the rate of
new developments or iterations in technology. For generative AI, this could be
the number of new algorithms or significant advancements in a given time
period.
2. Time to Regulation (R): This is a measure of the time
taken to draft, approve, and implement regulations related to the innovation.
This includes legislative procedures, public consultations, and enforcement
set-up.
3. Lag in Regulation (L): This is the difference between the
pace of innovation and the response time of regulation.
My Proposed Original Lag Formula:
\[ L = I - \frac{1}{R} \]
Explanation:
I (Innovation Rate): This would be quantified by innovations
per year or month (e.g., patents filed, research papers published, new
technologies released).
R (Regulation Time): Could be time units such as years,
reflecting the average duration from the onset of innovation to the enactment
of relevant regulations.
L (Lag): Represents the regulatory lag, with a positive L
indicating that innovation is outpacing regulation.
I today asked to compare the scores of this year vs. last
year based on the formula, and turned to Google Gemini for some AI assistance
to make a comparison. AI did the comparison, as requested but more
interestingly it provide a suggested revision to my formula, to deliver an even
more accurate result. It suggested adding an element of “C” to the
formula, based on the increasingly level of complexity of the technology that
must be regulated.
To calculate these scores, AI used real-world trends
observed in the AI and tech sector between early 2025 and today, February
13, 2026.
The Variables
I (Innovation Rate): In 2025, we saw major releases roughly
every month (I=10). In 2026, with the rise of autonomous agents and local
specialized models, the frequency of "disruptive" updates has more
than doubled (I=25).
R (Regulation Time): Governments have streamlined processes
slightly through "AI Safety Institutes," reducing the response time
from 2 years down to roughly 1.2 years.
C (Complexity): A new factor for 2026. Last year's tech was
mostly text/image generation (C=1.0 baseline). This year's tech involves
"Agentic Action" and "Autonomous Decision Making," which
are far harder to regulate (C=1.8).
1. Scores Using Your Original Formula
Formula: L = I - \frac{1}{R}
Time Period
Innovation (I)
Reg. Time (R)
Regulatory Lag (L)
|
Feb 2025
(Last Year) |
|
10 |
|
2.0 yrs |
|
9.50 |
|
Feb 2026
(Today) |
|
25 |
|
1.2 yrs |
|
24.17 |
Insight: Even though governments got faster at
regulating (R decreased), the sheer volume of innovation (I) caused the
"Lag Score" to increase by 154%.
2. Scores Using the Revised Formula
Formula: L = (I \cdot C) - \frac{1}{R}
This version accounts for the fact
that complex innovation is harder to regulate
than simple innovation.
Time Period
I
R
C (Complexity)
Regulatory Lag (L)
Feb 2025 (Last Year)
|
10 |
|
2.0 |
|
1.0
(Baseline) |
|
9.50 |
|
Feb 2026
(Today) |
|
25 |
|
1.2 |
|
1.8 (Agentic) |
|
44.17 |
Insight: When we account for complexity, the lag
score nearly quadruples (365% increase). This highlights that the
"gap" isn't just about the number of new apps, but
the nature of the tech. While R is improving, it cannot keep up with
the compounded effect of speed and complexity.
Summary of Comparison
|
Metric |
|
Original
Formula Lag |
|
Revised
Formula (with C) |
|
2025 Score |
|
9.50 |
|
9.50 |
|
2026 Score |
|
24.17 |
|
44.17 |
|
% Increase |
|
+154% Old Formula |
|
+365% Revused Formula |
|
|
AI’s Conclusion: My original formula effectively captured
the "Velocity Gap." However, the revised formula reveals a
"Complexity Trap": even as regulators become more efficient, the
increasing sophistication of technology (C) causes the regulatory lag to
accelerate at an exponential rate rather than a linear one.
To me I find this all fascinating but also frightening.
I will be adopting the revised formula and plan to run this assessment periodically (if the bots don’t get me first). I am generally a positive and upbeat person, as those who know me will attest. However, I want to share this information because I am concerned that the technology that I love and embrace has potential to cause great harm that we won’t be able to fix. The fact that quantum computing is advancing at a rapid pace as well only will serve to further compound this issue.
Glad I know a lot of brilliant people that are working on trying to close this gap in our ability to regulate technology, but it seems to me it is only going to continue to accelerate, despite all of our best collective efforts. Yet that won’t ever stop me from at least trying to help keep technology safe for us to use to our benefit.
Thoughts, on this formula? Thoughts on ways to help keep us
safe from new technologies’ impact on existing technologies? Any ideas to help
lawyers, regulators, or technologists learn more about these governance
challenges?
.png)
.png)
.png)
.png)
.png)
