Get your free personalized podcast brief

We scan new podcasts and send you the top 5 insights daily.

Jean-Denis describes Town as a token-allocation machine: it suggests where spending tokens can improve a user’s work. Cheap tokens make long-tail tasks worth doing, such as prep notes for every meeting or a monthly wholesaler price review that occasionally finds real savings.

Related Insights

The AI, Thomas, is given a daily token budget with the sole goal of maximizing the money it generates. The human team's only job is to improve the AI's underlying 'learning loop' to make this token-to-dollar conversion more efficient, not to direct its business strategy.

Current AI models are priced too cheaply, leading to inefficient consumption like using powerful models for simple tasks. As prices rise to reflect true costs, companies will need to optimize usage. This may create a new role, the 'Chief Token Officer,' responsible for allocating AI compute resources versus human capital.

Eric Glyman explains that per-token AI pricing fundamentally shifts software from fixed SaaS subscriptions into a variable marginal cost. Because token consumption behaves like rentable intelligence where autonomous agents subcontract tasks, managing AI spend increasingly mirrors managing headcount payroll rather than traditional IT licensing.

Microtransactions have historically failed because the 'mental load' of a human deciding on a small payment outweighs the value. AI agents, which can scrutinize tiny decisions without cognitive cost, can enable a new economy of per-use payments for data, content, and APIs.

Companies have moved through distinct phases of AI adoption: from ignoring costs ('token oblivious'), to gamifying usage with leaderboards ('token maximizing'), to a fearful cost-cutting phase ('token anxious'). The next, most effective stage is 'token smart,' focusing on spending wisely, not sparingly, to maximize value.

Howie Lu advises against anchoring AI costs to cheap software subscriptions. Instead, evaluate token costs against the opportunity cost of an equivalent human's time. A $150 agent-written board memo is cheap if it saves days of a CEO's time and produces a superior result.

To create truly proactive, helpful AI assistants (like a better Siri), systems must be willing to spend vast amounts of cheap compute "speculatively" in the background without a direct user prompt or a guaranteed return on that spend. This is only possible when the cost of a token approaches zero.

The initial approach to AI adoption was often "token maxing"—using as many tokens as possible under the assumption that more usage equals more value. A more sophisticated and sustainable strategy is "output maxing," which focuses on achieving the desired result while actively minimizing token consumption and cost.

Categorize all AI token spend into three buckets: 'Tokens that Teach' (valuable experimentation), 'Tokens that Produce' (work output), and 'Tokens that Spin' (wasteful, idle processes). The optimal strategy is to aggressively eliminate 'spin' tokens, optimize 'produce' tokens, and fiercely protect the budget for 'teach' tokens to foster innovation.

"Token Market Fit" is a concept to identify AI categories where high token spend is productive and valuable. Unlike traditional SaaS, where value is fixed, AI's value can scale with usage. Categories like coding and video generation, which support high token consumption, are seen as having strong "Token Market Fit."

AI assistants should be token allocation machines | RiffOn