Six months ago, building Tara Capital in its current form would have been extremely difficult.
Today, much of the platform has been built through a collaboration between human judgement and a team of AI systems.
That experience has taught me something important:
AI does not remove the need for management. It changes what management looks like.
As a former Army officer, portfolio manager and macro investor, I never expected to find myself managing a team of artificial intelligences.
I am not a software engineer.
Yet over the last few months, I have found myself doing exactly that.
ChatGPT became my strategist and sounding board.
Claude became my builder, helping with coding, website development and implementation.
Hermes became an operational agent running on a VPS, connected through Telegram and capable of producing research outputs and dashboards.
Individually, each system was useful.
Together, they became something more interesting.
They became a team.
The Manager's Job Remains
What surprised me most was that the role of manager did not disappear.
If anything, it became more important.
The AIs could generate ideas.
They could write code.
They could draft articles.
They could build websites.
But they still needed direction.
They still needed priorities.
And they still made mistakes.
One example occurred during the development of the Tara Capital website.
An AI generated social media contact details that looked plausible but were completely wrong.
When challenged, it admitted the mistake and explained that it had effectively guessed.
The incident was minor.
The lesson was not.
AI can be remarkably capable.
It can also be confidently wrong.
Human oversight remains essential.
Amplification, Not Replacement
Many people talk about AI as a replacement for labour.
That is not what I have experienced.
What I have experienced is amplification.
Tasks that once required specialist developers, designers, researchers and analysts can now be coordinated through a combination of AI systems and human judgement.
The human remains responsible for deciding what should be built.
The AIs help determine how to build it.
That distinction matters.
The Future of Small Organisations
Historically, building a research platform required employees, consultants, agencies and substantial capital.
Today, a small organisation can operate very differently.
A founder can work alongside AI systems that assist with coding, design, research, writing and analysis.
The barriers to entry are falling dramatically.
The importance of judgement is not.
In fact, judgement may become the scarcest resource of all.
What I Learned
The biggest lesson from building Tara Capital is that AI does not eliminate the need for experience.
It changes where experience is applied.
My years in portfolio management taught me how to evaluate evidence, challenge assumptions and make decisions under uncertainty.
Those skills turn out to be surprisingly useful when managing AI systems.
The technology is new.
The need for judgement is not.
Perhaps that is why I increasingly think of AI not as a tool, but as a workforce.
A workforce that never sleeps.
A workforce that can code, write, analyse and build.
A workforce that still needs management.
The future may not be humans versus AI.
It may be humans managing teams of AIs.