The Allocation Problem
AI doesn’t give you more time, it makes the allocation problem unavoidable
Monetisable Minutes
London’s in the midst of a 5 day tube strike as union’s push for a 32 hour work week. Once you’ve survived the Lime bike gauntlet, conversation inevitably turns to what these people are doing with the rest of their time?
As Sarah Tavel recently pointed out, perhaps the the biggest opportunity created by the mobile wave was the creation of more monetisable minutes. With screentime no longer confined to the desktop, online engagement exploded enabling the rise of the attention economy and a crop of socially isolated and anxious youngsters
Today, automation promises a similar shift. If AI creates leverage for workers, what happens to the minutes it frees up? As reasoning scales and prompts grow longer, the latency itself becomes a perfect moment to check Twitter or reply to the group chat. Anecdotally I have a friend in PE who logs 24hrs Tik Tok screentime a week! Left unguided, AI productivity might create more monetisable minutes rather than better business outcomes
The immediate challenge for start-ups and their customers lies in attribution. How do we measure whether freed-up minutes translate into genuine productivity rather than just more activity? Enterprises experimenting with Co-pilot and ChatGPT licenses often default to surveys asking how much time was saved. That ‘vibe check’ may be directionally useful, but it’s not enough. As Coinbase has highlighted, organisations need to be intentional: set objectives, capture baselines, and methodically track outcomes. This new walk creates opportunity for start-ups like Paid to lay the enabling infrastructure
Now in some sectors I do believe we will see the opposite effect, a kind of Jevons paradox. Instead of eliminating work, AI multiplies it, raising the bar for quality and creating a Sisyphean struggle to keep up. This mirrors what happened in law and finance, where computers improved efficiency but simultaneously pushed expectations higher. 9-9-6, the pursuit of the 1-person billion dollar company and locking in to escape the permanent underclass are precursors of this vibe and a future where the gulf between ‘the best’ and the rest grows dramatically
The Human Edge
Efficiency on its own doesn’t guarantee growth and without deeper intelligence on what to prioritise and when to act you risk creating a tool for speed rather than strategy. Worse, those freed minutes may simply flow into someone else’s P&L
This is why we need systems of action, platforms that go beyond static databases and instead actively guide workflows, automate processes, and direct humans toward tasks where they still command a skill premium (original thought, human relationships etc.). In a world where we’re all 10x more powerful, the hard part is deciding what’s worth doing
One segment where this dynamic is especially clear are brokers. Their back-office work is ripe for automation, client advisory augmentable but their sustaining value comes from human leverage - trust, relationships, and judgment in complex markets. AI can help augment their core functions around market access, price discovery and decision support creating opportunities to either rebuild tech-enabled brokerages or empower the individual via tools that originate more deals, sharpen pricing, and free up brokers to deepen trust:
And the implications extend beyond classic brokers. Any service where value is anchored in human relationships could be reshaped, i.e. professions requiring a credentialed human and/or contexts demanding high trust. These opportunities are particularly interesting if you can help shape customer demand, reduce a major shared COGS with scale and ride or catalyse supply of new entrants:
AI doesn’t give you more time, it makes the allocation problem unavoidable. Start-ups that matter won’t just hand people back minutes, they’ll tell them where to use them
If you’re tinkering around this theme I’d love to chat - jamie@triplepoint.vc