Antonym: The Two Lawyers Edition
Dear Reader,
As we often point out, generative artificial intelligence (gen AI) is a cognitive accelerator – a turbocharger for the brain. That sounds exciting, right? Who wouldn’t want to think faster?
But this week, we’re taking a step back to consider: what happens when acceleration has consequences? We explore a thought experiment involving AI-powered lawyers in a modern twist on the fable of the tortoise and the hare. We also think about how teams must adapt when everyone’s cognitive engines are running faster than ever before.
Speed isn't an unalloyed good. Just as vehicles need speed limits and even the swiftest runners face injuries, there are moments when slowing down can be valuable – consider the virtues of slow food or slow travel. Generative AI supercharges cognitive processes, but it's worth acknowledging that faster isn’t always better.
Sliding lawyers
Let’s try a thought experiment: Picture two lawyers' offices in London's legal district. In one, a gifted litigator who's always been the smartest person in the room. In the other, a steady performer who's built a respectable if unremarkable career. Both are about to be transformed by artificial intelligence – but not in the way you might expect.
The conventional wisdom about AI in professional services has been straightforward: it's a tool that amplifies existing talent. The better the lawyer, the better the results with AI. But emerging evidence suggests a more complex and fascinating reality.
Sometimes AI does not just amplify capabilities – it rewards learning agility and persistent innovation more than raw talent.
Part 1: Closing the Gap
Initially, both our lawyers embrace basic AI tools. The naturally gifted lawyer quickly masters document review algorithms and contract analysis software. Their already impressive work becomes even better. The average performer takes longer but achieves something remarkable: their work quality begins approaching that of their more talented peer.
It's like giving everyone a calculator. The person who's brilliant at mental maths still has an edge in understanding what calculations to make, but the basic operations become equally accessible to all.
Part 2: The Plateau and the Climb
Here's where it gets interesting. Our brilliant lawyer, satisfied with significant initial gains, sees AI as a useful addition to their toolkit. But our average performer, energised by newfound capabilities, dives deeper.
They build custom algorithms for predicting judicial outcomes. They create client-specific chatbots that handle routine queries. They develop pricing models that align their firm's interests with client success. What started as catching up becomes breaking new ground.
Often, those who've succeeded through traditional means have less incentive to fundamentally reimagine their approach. It's the professionals who've had to work harder to compete who are more likely to embrace transformative change.
This phenomenon isn't limited to legal practice. Across finance, medicine, and consulting, similar patterns are emerging. The real competitive advantage may not be raw intellectual horsepower but the ability to continuously evolve with technology.
The implications are profound. As AI becomes more sophisticated, the gap between natural talent and acquired expertise might narrow further. The key differentiator could become not how brilliant you are, but how effectively you can partner with AI to reimagine your profession.
We're moving from a world where being the smartest person in the room was enough to one where being the most adaptable and innovative is what matters.
For law firms, corporations, and professional services organisations, this suggests a necessary shift in hiring and development strategies. Raw talent remains valuable, but the capacity for continuous learning and technological innovation may prove even more crucial.
This thought experiment could run in different ways. Give it a go yourself.
I’m reminded of the 2022 paper “Artificial intelligence and the changing sources of competitive advantage”, which looked at how chess tournaments that allowed humans to compete with machines and hybrid AI-human teams (known as “centaurs”) played out. Generally, the “centaur” teams – people plus AI – beat the best AIs and the best humans on their own.
Even more interesting was that the winning centaurs weren’t comprised of the best chess players from the old system, but players who learned to work with AI and data best to inform their decisions.
In the paper’s words:
[…] AI adoption triggers interrelated substitution and complementation dynamics, which make humans’ traditional competitive capabilities obsolete, while creating new sources of persistent heterogeneity when humans interact with chess engines. These novel human-machine capabilities are unrelated, or even negatively related, to traditional capabilities.
As our imaginary average lawyer demonstrates, in the age of AI, the race doesn't always go to the swift – it goes to those who never stop learning how to run better.
AI Power Hour publicity
This week we launched our “smallest ever consultancy product” at Brilliant Noise, the AI Power Hour. It got one of the strongest responses of any post we’d put on LinkedIn before, which shouldn’t have surprised me, but did. The concept and the offer seemed to resonate very well, so I’m sharing it here too.
Experience AI in Action: In just 60 minutes, tackle your project – whether it's a report, research, or business plan – and make tangible progress, often completing it entirely.
Transform How You Work: See how AI organises, enhances, and accelerates your workflow, turning scattered ideas into structured insights.
Support a Good Cause: Book before year-end, and for every session (£300), we’ll donate one to a charity.
If you want to book your own AI Power Hour, reply to this newsletter to book your AI Power Hour today.
Z-Pain will pass
The rule of Mark Zuckerberg is that 7/8 of the zuck is below the surface. That awesome floating mountain is only an eighth of the actual whole mass of zuck. Occasionally, because of a warm current or general entropy the Zuckerberg tilts to one side and we can see some of the usually sub-surface spectacle.
This week Mark Zuckerberg released a song on Spotify, a cover version collaboration with T-Pain of his classic “Get Low” (TBH I don’t remember it.)
In a week where billionaires-doing-whatever-they-like involves things like setting up US government departments with joke names, Zuckerberg’s duckery seems almost endearingly harmless where in most timelines it would be an affront to the ears and decency.
For comparison, here’s the original track:
Sample lyrics:
Get low
To the window
To the wall
Till the sweat drop down my balls
Till all these bitches crawl
And here’s the new version. Zuckerberg sings rather than raps, and plays along on an acoustic guitar.
Aww, sweet. I hope T-Pain got buy-a-president levels of fees for his service.
Musk Apology
There was an error last week’s Antonym. Thank you to sharp-eyed readers who spotted it.
The value (market capitalisation) of Elon Musk’s companies did not increase by $15BN following his US$119 BN as originally stated in Antonym: The BROTUS Edition. It was much more.
Elon Musk’s net worth has increased by approximately $70 BN, reaching over $300 for the first time in nearly three years since Donald Trump’s election victory on November 6th. This increase is largely attributed to a 39% rise in Tesla’s stock price following the election.
Illuminate
Back to feel-better news.
Google updated Gemini, its rival to ChatGPT, this week. It’s very powerful, but also very prissy. If you can do anything without it accusing you of trying to trick it into doing something against its apparently very strict guidelines on acceptable use, then you are very lucky. Frankly, it gets in the way of its usefulness.
Meanwhile, it’s still a love-fest out there for the incredible podcast-generating abilities of Google’s still-experimental, free-for-now NotebookLM. If you haven’t seen its academic-inflected version, Illuminate, I highly recommend taking a look. It’s less frothy in its tone than the enthusiastic NotebookLM voices.
I particularly like that you can listen to a podcast generated to explain Attention Is All You Need, the paper from eight Google computer scientists that basically invented the Transformer (the “T” in “GPT”). The original paper it is too technical for most of us mortals to unerstand, but the podcast does a great job of explaining what it says.
This week’s…
Reading
Tomorrow and Tomorrow and Tomorrow, by Gabrielle Zevin
This a book that carefully pulls your attention in its direction, and then you find you can’t look away. So warm and full of longing. I’ve got the physical book but I started listening to the audio-book and now I can’t stop.
Creation Lake, by Rachel Kushner
Short-listed for the Booker Prize, Creation Lake is about an undercover agent infiltrating a French environmentalist group. I will turn up for Kushner’s prose whatever she’s writing. I love the wildness of the plot as well.
Watching
Anora (Cinemas)
This movie is never what you think it is going to be. The plot is high concept: a sex worker falls in love with a the son of a Russian oligarch while he is visiting New York. When they get married on impulse, his parents are furious and send heavies to intervene. Things go awry and… hilarity ensues? Well, farce and tragedy and a chase movie. Highly recommend.
The Mirror & The Light (BBC)
My love for Wolf Hall books and TV series, for Mark Rylance and the whole cast is very big indeed. This series is wonderful and has prompted a complete re-watch of the first in our house.
It also reminds me of the closing line of Bring Up the Bodies, the second book of the Wold Hall trilogy:
The word ‘however’ is like an imp coiled beneath your chair. It induces ink to form words you have not yet seen, and lines to march across the page and overshoot the margin. There are no endings. If you think so you are deceived as to their nature. They are all beginnings. Here is one.
That’s all for this week…
Thank you for reading. Click on the 🤍 below to let me know if you liked it.
Antony