In late 2024 I tested ChatGPT 4o for contract review. Generally it summarized well and it caught some things that I missed and it missed some things that I caught. It saved me a bit of time, but it didn’t make me rethink my career. I filed it under useful tools and carried on.
The model releases kept iterating and after a few months Claude Opus 4 arrived quickly followed by GPT 5. These models made the 2024 models look pedestrian, both in terms of intelligence and model reasoning. The rate at which they were improving amazed me. I realized that if I wasn't using the latest models I wouldn’t be able to see where adoption is heading.
In January 2026 I setup OpenClaw, an autonomous AI agent running on my own computer. Two things about it landed differently. The first was that it had real autonomy, in that I could give it instructions and it would go away and do things on its own. The second was that I built the thing myself, on my own home computer. That second part is what got me. As a kid I was a tinkerer and my idea of fun was to build my own computer - connecting hard drives, upgrading video cards, installing additional ram, building something useful for myself. And so now I am reunited with my inner tinkerer and with the AI augmentation, I am able to produce things that I wouldn't have touched just a year ago. That's the shift I'm highlighting, being out of my depth no longer stops me so long as I am persistent.
With the speed at which AI is moving, it was obvious to me that that one of the types of work that I did previously - contract drafting, review and negotiation, would ultimately be performed by AI at some point in the future. That’s not to say contracts lawyers don’t have a future. Far from it, in fact I believe Jevons paradox applies here and we will need more contracts and more lawyers as the cost of producing contracts falls due to AI. However, it did make me examine my own career and, to be honest, I didn't love the law enough to continue working as a lawyer. This is why I've left my prior career to pursue AI full time, focusing on practical AI operations for small professional services firms.
If a small law firm had virtually unlimited employees how would it operate and do business? What would a small law firm operating system look like? I believe the future evolution is AI agents taking care of firm operations and lower level legal work while human lawyers reserve their time for higher level legal judgement, business strategy and human-to-human client servicing. The question is how do we get there? The problem with AI is that it's like drinking from a firehose and so it’s difficult to know what the next step for small firms is in the AI adoption journey. I plan to post here what i learn and so that we can all explore together, because the real answer at this moment is that nobody really knows.
So honestly, you don't actually need a strategy, a budget or a plan in order to adopt AI. Take one workstream that you do every week that you have never enjoyed, and spend an afternoon seeing how far you can get integrating AI for that specific type of work. If that afternoon goes well, you'll want the next one.