March 2025 Updates

Ai, betting & biggest opportunities continued

Late Feb and into March seems to be another massive leap forward, it has the same vibe as what happened in late October updates for the utilization of Ai in business.

This month we built a new Amazon software tool including authentication, back-end-operations, UX and deployment in 5 days. This could have been reduced to about 20 hours if I had any idea what I was doing with any of this but about 45-50 hours in the end.

This was basically just to learn how to build something and how far things have come from about a year ago, this project would have cost 5 figures and taken a few months (at least), now for a couple of hundred and 5 days it was completed.

The site is here for anyone interested: https://amzlens.com/ - it’s a simple build but surprisingly difficult in the weeds (for a non dev).

Aside from this the Claude 3.7 extended edition update has continued to be even more of a game-changer.

As big a step forward as the October Claude 3.5 update for my personal projects.

This has enabled the sports betting models to get sharpened up rapidly, including the slower tasks that usually prevent quick growth (scraping custom data, parsing into JSON etc).

With these mega-efficient updates, all the models across the board are adding another 1-3% ROI (per bet) now is making this the biggest opportunity by far.

Results are starting to align to back-tests as well. Some sports actually running at 20% ROI per bet, which is just wild to think about, don’t think that’ll last forever but anything over 10% at scale is incredible results.

The goal moving forward now is to move towards our £50k/month profit goal structure across the board and work out exactly how this would look in the quieter summer months when there’s a lot less volume of what we bet (rugby union, hockey etc).

We’re at the beginning

One thing I’ve been talking about a lot recently is how we’re still at the beginning of all this micro-ai-saas/cool concept builds related to AI.

Even though if you are on YouTube or social media alot and your algorithm is showing you every kid building 6 and 7 figure Ai tools in 10 minutes it seems like we’re behind.

This is so wrong.

It’s the start of the next goldrush for people who have the drive to build and build quick.

Previously getting “MVP’s” built in a few months and just launching to see what the market says was the play.

Now you can build MVP’s that are 100 times better in 1/10th of the time, the only thing missing is drive and ambition to do it quickly and well.

Things still break but you can get apps or saas or models built in days not months now and they can earn for you forever.

Similar to the original post on this site, build something valuable once and earn indefinitely.