PUT THE HUMAN BACK INTO THE WORKER

Solo Founder (AI)

MEDVi

I saw an article in the NY Times reported by Erin Griffith from Los Angeles about a solo founder, Matthew Gallagher along with his younger brother Elliot, who spent two months acquiring more than a dozen AI Agents by spending $20,000. He started a company called MEDVi. AI created a website and wrote the code that powers this company. AI also generated images and videos for ads, and handled customer service. There is a human in the loop that can be contacted. This is an important consideration because the government wanted to contact that human in the loop. MEDVi engaged in some outsourcing when needed for stuff the two human employees didn't want to do themselves in the short term. MEDVi is a telehealth provider of GLP-1 weight-loss drugs that it sells cheaply, and delivers without requiring a visit to a doctor.

Innovative and Scalable Still Needs Oversight

I'm not exactly sure how MEDVi acquires the GLP-1 weight-loss drugs that are distributed. However MEDVi's aggressive advertising created some regulatory issues. In February the FDA sent it a warning letter about false/misleading info on the MEDVi website that suggests MEDVi mixes its own FDA-approved drugs. As far as I can tell they don't. This is a possible misbranding and, therefore, MEDVi exists at the crossroads of incredible innovation and scaling, and ongoing regulatory challenges. The MEDVi.org site shows the company is still in business. As far as I can tell the founder hasn't violated any laws. I suspect the FDA is not serious enough (so far) to ensure they are behaving correctly. They seem happy enough if MEDVi just corrects problems after they are reported. This feels to me like the government's strategy is to just check and see if the problem is eventually corrected, or at least that there is a claim it is corrected. So many government departments, nowadays, lack the staff they need to do a thorough investigation into reported problems and then do follow ups as needed.

For the most part, the government's actions tell me that it loves startups and seems willing to give them a very long leash if questionable behavior is detected as long as they are generating enough sales and profits. In the past sales and profits were used to pay the salaries of employees. The existence of employees also gave the government some incentive to work with companies because it also helped the workers, which also helped the overall economy. The current trends are pointing in a different direction. Medvi has only two employees in this entire company. I believe it helps dampen the government's risk investigation if a startup has more than a billion dollars in sales, or appears to be headed in that direction. I mean maybe, someday, they will find ways to feed some of that money to the government or whatever administration is currently in control. Perhaps they can help privately fund a ballroom.

Billionaires learned long ago that they are mostly untouchable. They can just largely ignore the government cause the billionaires fund the campaigns of politicians. Therefore, one hope startups often have is to create one or more billionaires along the way. If human employees are never there or are eliminated, the first billion dollars comes much more quickly. I can only imagine the dreams one can have of owning a walled-off bunker with your own robot security firm protecting you from the hordes of poverty stricken masses roaming around outside wishing they had enough food.

What This Means For Everyday Workers

I can't help myself, First I absorb what is happening right now, and then I start looking for longer term effects. Where is this heading? What can we do? So, I think this "as few employees as possible" strategy is the start of a growing movement in company creation. Will the demand grow for access to knowledge about controlling AI Agents that take on tasks that could be done, or are currently being done, by humans? The open-source project 'OpenClaw' can be used today to create AI Agents, and these AI Agents can even create and control other AI Agents if desired. AI tools can be given to these agents that allow them to access actions that affect the company's entire computer network.

For example, consider the scenario where almost all workers are AI Agents. These AI Agents are first used to create a company that does people's taxes cheaply. This company is given the name MyOwnTax. The AI Agents can create a website, advertising, software to calculate taxes owed or refunds expected, and processes to handle customer service. Even if problems are detected with this new company, say with refunds or the amount owed to the IRS, it's likely the IRS, after the recent massive layoffs of IRS human employees, no longer has enough resources to come after this new company with much force. They'll just ask that the problems be fixed in a good way. The company might make billions of dollars because expenses are low and the demand is high. Everyone will be so impressed by the speed and scale of growth. MyOwnTax will not have to worry about paying much in salaries or healthcare because the vast majority of workers are AI Agents. No one will be able to touch any newly minted billionaires.

Does this trend of fewer human workers and more AI Agent workers help society? I believe it does the opposite. Instead of improving communities, it contributes to societal deterioration by contributing to the threat of eliminating 20% of all desk jobs soon (at least according to a recent NY Times editorial), along with the rising cost of healthcare, housing, education, and food. This motivates me to promote campaign strategies that focus on increasing wages so voters can afford better lifestyles and live in nicer communities. If this strategy gets the right people elected, then we can start working on legislation and regulations that address affordability, wealth concentration, and improving communities (e.g., rural hospitals need help). Maybe more humans can be hired to better investigate potential risks.

Human Workers Are Valuable

Right now underemployment for new grads just reached 43% and job openings are the lowest since 2020, making this a bad hiring market (Bloomberg). The current vibe is referred to as a 'techsession' because computer programming jobs have declined in the U.S. to their lowest levels since 1980 (before the Internet existed). College graduates, Ages 22-25 saw employment fall 20% since 2024 (FED). A bad entry-level job market means many graduates will not be given the opportunity to gain as much experience over time. This can be bad for companies that are growing and need experienced employees to carry on with needed work after senior workers retire or move on (Bloomberg). Many Tech CEOs claim jobs will disappear. The CEOs of Anthropic and Nvidia have set the limits of the range of possibilities: Anthropic'c CEO warns that AI will cut half of today's jobs, while Nvidia's CEO thinks it will produce more jobs (Vanderbilt Policy Accelerator: "After the AI Crash").

The Government Share (AI)

I haven't been able to figure out how AI will contribute to an increase in Tech job creation, but temporary increases in blue color jobs for electricians and plumbers is believable. It seems prudent to think about the possible impacts if there are a large number of job loses in other areas. I am thinking about ways to encourage job creation. I was reading that in August of 2025 the Trump Administration told Nvidia and AMD that export controls on semiconductor sales to China would be lifted if 15% of the sales went to the Government (I'm assuming it doesn't go directly to President Trump). Then in December of 2025 it did it again only this time it was 25% of sales of advanced semiconductors. This feels unconstitutional to me and I'd be curious what precedents there are for this type of economic intervention.

See https://www.cato.org/free-society/winter-2026/conspicuous-fist-trumps-state-corporatism# for more details.

I'm not exactly sure what a "golden share" is, but I'm pretty sure it gives the government some more control over some decisions. This all feels unconstitutional to me (but what do I know).

As long as we now have some strong precedent here, let's look forward and ask, why not let the government raise even more money by focusing on the most valuable Tech companies? Promoting the expansion of this practice into other areas will raise even more money that can be used to retrain those affected by layoffs and recent college graduates searching for ways to start a career. What an interesting idea that could also be used to help reverse the flow of money that contributes to the concentration of wealth into the hands of few billionaires. What about including the number of humans a company employs as input into the calculations used to compute the acquisition percentage of sales. If a company employs many humans making incomes at or above the median value, then their acquisition percentage can be lowered. If a start-up company has NO HUMAN EMPLOYEES it must pay a large percentage of all sales. I need to do more research here though. Defining the formula used to calculate the percentage of sales the government acquires is complicated. The idea is, if a company's business strategy focuses on reducing human employees while increasing the number of working AI Agents, then the government should have a say about what happens to the profits being generated. Tech companies like Nvidia have already acquiesced to this coercion by the Trump Administration. There was no national emergency but it was done anyway. All the Administration had to do was threaten them with denial of foreign exports. Nvidia caved because they just wanted to make sales to China. Feels like an opportunity to me and I suspect voters will agree, and voters still have power.