264 | Whatever You’re Doing With AI — Pause + OpenAI leaks new feature

Hello to all 7820+ of you from around the globe. It’s been a minute.

After 263 newsletters, I stopped writing. Not out of fatigue — but out of respect. Respect for you. Respect for my own reckoning. Respect for the struggle to find signal in the noise — the kind that demands we ask:

What exactly are we building with AI?

And just as urgently: What is AI building (or destroying) in us?

Over the past two and a half months, I’ve been sitting with those questions. Watching the noise grow louder, the claims bolder, the discourse thinner. And in that fog, I didn’t want to be one more voice guessing at the future or branding themselves as the latest AI whisperer. I wanted to wait until I had something worth saying. Something that cut through it.

So here it is.

The Reversal We Didn’t Expect

When I first began exploring large language models in legal practice, I was among the earliest voices talking about their potential to reduce cognitive friction — especially in learning. Whether you’re a 1L student or a seasoned partner learning a area of law, there’s real pain in the learning curve. That pain is what actually generates the real learning. And AI seemed poised to help.

• A tireless tutor.

• A responsive assistant.

• A clarity engine for complex ideas and topics.

But now, as I feared, I’m noticing something quieter and more troubling. Not the automation of tasks — but the erosion of thinking. The outsourcing not of labor, but of learning. We now have tools that can summarize, respond, and even argue persuasively. But they do so without understanding anything at all. Fluent? Maybe. Fast? Absolutely. But fundamentally disconnected from comprehension, context, or consequence.

That’s not intelligence. It’s what one writer recently called anti-intelligence — and it stopped me cold. Anti-intelligence is not stupidity. But an inversion of knowing.

A convincing simulation that makes it harder, not easier, to recognize what actual intelligence is. Think of how many stories you have heard where people are believing that their conversation with AI is real – as in the AI “knows” them and is “speaking” to them. They forget it is all probabilistic configurations of language, not true connection or understanding.

Fluency ≠ Understanding

And this is where the problem gets deeper — and more personal.

Because if we mistake fluency for understanding, or speed for insight, we’re not just using a new tool. We’re reshaping our expectations of what it means to know something.

Over the past year, I’ve noticed the questions shifting.

From: “How can this help me think?”

To: “How can I keep up?”

That is dangerous. And in that shift, we’re letting go of the very things that make legal reasoning valuable: Judgment. Doubt. Interpretation. Time. Nuance.

We are not here to be fast.

We are here to be right, and responsible, and clear. 

Reclaiming the Pause

So, here’s my invitation to you — especially those of you who feel the speed and AI impact creeping into your brain and your work:

Pause.

Put down the prompt.

And ask yourself:

• Am I using AI to accelerate my thinking — or to avoid it?

• Is this making me learn — or just helping me produce more?

• Am I thinking better — or simply thinking less?

Regular readers know that I’m not anti-AI.

I’m anti-hype.

And I’m pro-human.

Let’s not mistake output for authorship.

Let’s not confuse coherence with care.

As Luck Would Have It…

As I was working through this — creatively stalled, mentally jammed — something curious happened.

In the last 48 hours, leaks began circulating about a new feature being tested by OpenAI inside ChatGPT. It’s called Study Together. Not much is known about it yet, but here’s what we do know:

It shifts ChatGPT into a more tutor-like mode — asking questions, walking users step-by-step through concepts, and guiding rather than just answering. It feels less like a search engine, more like a Socratic coach. 

The target audience? Likely students. Possibly classrooms.

The business angle? Probably another upsell or educational push.

But the implication? That matters a lot more.

This is OpenAI — perhaps unintentionally — admitting that something essential has been missing. That in all the fluency and function, we’ve been stripping away the substrate of thought.

I’ve written about this before: the importance of cognitive friction. That moment where your brain resists, strains, and learns. Without it, we’re not gaining insight. We’re just downloading performance.

You Don’t Need a Feature to Think This Way

Whether or not this new mode rolls out widely, here’s the good news:

You don’t need “Study Together” to study better.

You can make any AI tool your guide — instead of your ghostwriter — by simply changing how you engage:

Try this prompt next time you’re studying a concept, preparing an argument, or learning a new area:

“Act as a tutor. Don’t give me the answer. Ask me questions step by step. Let me try. Correct me when I’m off. Help me understand — not just repeat.”

This is what I’ve always advocated:

• AI as a simulator. A scaffold. A mirror.

• Not to remove the thinking — but to help us stay in it longer.

And in my adivsory and acdemic work with law schools and firms, this very idea — of building AI mentors and tutors — is becoming a key area of focus. How do we preserve the discipline of learning with AI, not just because of it?

Where I Go From Here

I took a break from writing not because I ran out of things to say — but because I didn’t want to say things that didn’t matter.

But now, it feels necessary again. Not to chase the latest update or prompt hack, but to remember what matters most in this moment of transition.

In the coming issues, I’ll return to more grounded explorations of AI’s role in legal practice — not as magic, but as craft.

Not as shortcut, but as supplement.

Not as the answer, but as a better way to ask questions.

Thanks for reading — and for sticking with me through the pause.

To read previous editions, click here.

DISCLAIMER: None of this is legal advice. This newsletter is strictly educational and is not legal advice or a solicitation to buy or sell any assets or to make any legal decisions. Please /be careful and do your own research.