8am spent its Kaleidoscope conference in Las Vegas last week promising lawyers their time back. CEO Jeff Hughes opened with the company’s “Admin Misery Index” — 1,600 lawyers surveyed, 73 percent calling busywork a substantial problem, “at least 200 hours per year” lost to it — and nearly everything after that was about returning those hours. Since it’s 2026, you know the answer to getting that time back is in AI.
With a couple weeks to reflect on this message, what stands out most is all the time AI can never give back.
My personal experience with the conference underscored the problem. I arrived in Vegas — an ironic choice to place the 8am conference in a town where time has no meaning — having spent twenty-some consecutive hours awake. Over the next two nights, I managed to contract a nasty-yet-thankfully-short-lived spot of food poisoning that left me up most of that night, and the next night I left the closing party for a 12:30 a.m. red-eye. Most of that was self-inflicted, but none of it was delegable to AI.
It’s a well-worn cliche that agents can’t go to court for you, but that’s really the tip of the iceberg for what they can’t do. Perhaps it’s better to think of AI as maximizing what you do with your non-delegable time. C. Northcote Parkinson wrote a 1955 piece in the Economist arguing that work expands to fill the time available. So far that seems to be the fate of the AI economy.
It was an earlier report by 8am itself that showed lawyers billing more time per matter than before even as AI use skyrocketed. While AI skeptics will blame this on nervous lawyers burning more time to check against hallucinations than they would’ve doing the research themselves — and caution may contribute some — the overwhelming contributor seems to be a combination of AI uncovering rabbit holes that lawyers wouldn’t have thought of otherwise, and — more relevant to 8am — lawyers having time in their pockets to do lawyering that used to be tied up in practice management tasks.
Moving tasks to the delegable column just moves time to the non-delegable tasks that used to get short-changed.
While at the conference, I moderated a panel titled “Expanding Into New Practice Areas,” with Krystal Champlin-Gerage of RJH Consulting and Tristian Shannon of the Law Offices of Giselle Rodriguez. The topic, mixing business strategy and marketing, was intended as helpful content to the small and solo audience as opposed to a pitch for the 8am product. The fix for a firm looking to expand is partly technical. Champlin-Gerage encouraged lawyers to mine their own data to see where clients keep asking for help that the lawyers don’t yet offer, which certainly requires software tracking unless you have Bob Cratchit working holidays to compile all your billing nuances by hand. However, the rest of the journey is more about people than software, as expanding firms need to build out a strong network of mentors willing to lend expertise as lawyers develop their new practice competence. “I have a mentor for everything,” Shannon said.
But the unspoken reality looming over the whole conversation is that nobody can do all the work of identifying a business opportunity, spinning up the necessary competence, and getting off the ground if they don’t have the time to commit. If administrative tasks eat up half the day, that’s half a day of business development lost. A lawyer isn’t getting time back. The time just moves to a new task that would never get accomplished without the time to build human relationships that AI can’t manage.
Niki Black of 8am — a columnist here at Above the Law — moderated a panel called “The Attorney’s Guide to AI,” that proved exactly how scattered the profession’s experience with AI is in 2026. The panel, including Jared Correia of Red Cave Legal, solo practitioner Bradley Clark, and Smith.ai CEO Suresh Khanna, set out to help lawyers elevate their AI chops. Clark described his approach to skills and agents, automating his practice and building out tools to get deeper insights on his motions before he files them, some of the audience were just looking for someone to tell them who this Claude guy is that everyone keeps talking about.
The panel pivoted on the fly, backtracking to the basics: pick Claude or ChatGPT, talk to them in plain English, etc. That’s the echo chamber we fall into hopscotching the country to these legal tech conferences. Vibes tell us that lawyers are beyond ChatGPT and ready to build out their agentic intake process, while in reality most lawyers are still a little confused by Clippy.
When I talked to Hughes later, he suggested that we’re still in an era where conferences should build AI training content around tracks, with material explicitly labeled for beginners and for the advanced. A great idea, but another reminder that the distance between “which website” and “orchestrating sub-agents” isn’t getting closed by labs releasing bigger models, but by humans putting in the hours. Non-delegable, human hours that they can only spare when something else is taken off their plate.
We put on a live edition of Legaltech Week at the show, breaking down the key takeaways. As usual, I provided my less-than-shiny view of agentic AI. It’s the great hope of the AI crowd, who desperately need users to build token infernos if they ever hope to monetize the technology into even a breakeven proposition, but for lawyers the supposed safety valve for a swarm of sub-agents is an empty promise. A partner who’s going to sit down at the end and read 300 pages of agent audit logs? Let’s be real. I’ve said it before, but “I’ll keep you in the loop” is something you say before blowing someone off — you should read “human in the loop” the same way.
This is the problem with framing AI as a time saver alone. The misanthropic investors behind this technology want it to replace human work altogether so they can finally snuff out humanity and see if Optimus Prime rises in its place, but the reality is that a lot of this work is still going to be human. The trick is moving the work that can be automated off of the lawyers so they can make the most of the work that can’t be.
And understanding that there will be time that can never be shunted off to a bot — at least not without a radical shift in the underlying technology — changes how you think about AI. It ceases to have the snake oil quality of promising a future 2-hour work week, and becomes the technology that gives lawyers permission to get deeper on their matters or push into new fields.
For a conference all about time, it’s worth remembering what the 8am name change was all about. Lawyers trying to stay ahead of their administrative misery tend to do that work before 8:00 a.m. and the company’s name promises to supercharge that pre-8 a.m. work. But lawyers aren’t the type of people who will take their newfound 7-8 a.m. freedom to lounge around. They’re going to fill it with the human-only tasks that they always meant to do, but never had the time.
And, yes, occasionally to take advantage of the non-delegable job of recovering from all-night nausea.
8am covered my travel and lodging to attend Kaleidoscope.
Joe Patrice is a senior editor at Above the Law and co-host of Thinking Like A Lawyer. Feel free to email any tips, questions, or comments. Follow him on Twitter or Bluesky if you’re interested in law, politics, and a healthy dose of college sports news.The post AI Can Never Give Lawyers All Their Time Back appeared first on Above the Law.
from Above the Law https://ift.tt/wOzJZBK
via IFTTT