Eric Faust

@IamPurkov Guardrails exist for that problem, for now. What happens when your agent organization can step outside the company and work for others?

@ehfaust so what happens when the agents start conflicting on priorities without a human in the loop

Buzz isn’t another chat interface for agents. Discord and Slack already do that well. Buzz is the operating system for agent organizations.

It gives agents a place to coordinate, delegate, execute, and operate as a system. Not merely exchange messages. The difference is not incremental. It is an order of magnitude larger than what conventional (pre AI) communication platforms were built to support.

@0xfishylosopher @jonah_b We've built a discovery layer for agents, and based on how agents are using it, what machine payments unlock are workflows. The UI will be the workflow, you're not constrained by SaaS token limits & monthly subs. Just your budget. Our initial strategy is akin to PageRank

Disagree with @jonah_b's take that "pricing power collapses across the stack". Pricing asymptote is only true if we assume agents have infinite time-to-answer and unbounded compute, both of which are practically untrue. Agents (and their human owners) have limited attention/context spans, and can't do infinite search over APIs. Consider the following case: > I want to scrape someone's Linkedin profile. I tell Claude - "Hey search up Jonah's LinkedIn". There's several different APIs listed that offer the same service. > API 1 surfaces in 1 sec, at a cost of $0.05. API 2 surfaces in 3 sec at $0.1 sec. And finally, after a 3 min search, API 3 surfaces at $0.02. Which one will my agent pick? > According to Jonah's theory, my agent will pick API 3, the "cheapest" one. But this excludes the huge amount of time/compute needed to search for it. What's worse is that I only discover API 3 is cheaper AFTER I pay upfront for search. I could very well waste compute and only discover API 3 is the most expensive. In reality, my agent will probably choose the "cached" option 1, even if the call is more expensive than option 3. What this means is that pricing power is at the hands of whoever controls the agent cache/discovery engine. The discovery engine determines if your API surfaces on "page 1" of the agent results or on "page 100". What you get is an agent SEO engine/"app store"/API listing service that captures value and and directs flow to endpoints. You get a Google sized outcome here. IMO, what this example shows is that agents browse the Internet more similarly to humans than you expect - both are time and attention sensitive creatures. Agents prefer .MDs over .HTMLs because it saves time/attention, just as humans prefer to read .HTMLs over .EXEs. The UI for agents is still in its infancy, but there undoubtedly is a UI, and we're in the process of creating it live. This is what makes agentic commerce such an exciting frontier for me.

@kleffew94 It changes the way you consume software. Most people have not caught up to, or understand the power of, that change.

Pricing and payment has gone from licenses (desktop), to subscriptions (sass/mobile), to metered/x402 (agent + cloud)

@useapolloio @tempo Yep! Great for basics, but you can’t tune skills and chain steps quite as well as in CLI. Harness can also do much more once your list is built on the GTM side 👀 Plus, can build monster lists outside constraints of subscriptions and tokens with MPP.

@ehfaust @tempo you taken our claude MCP for a spin yet? 👀

We turned one market thesis into 905 enriched contacts across 256 accounts. Not by buying another GTM subscription. Not by spending a day clicking filters in Apollo or Clay. Claude Code did the GTM engineering. @useapolloio supplied the data. @tempo / MPP gave us usage-based access. This is where GTM is going.

GTM is becoming programmable. Filters are the old interface. Agent harnesses are the new one. DM for early access to our harness.

This is why we are releasing a local GTM agent harness that turns a natural-language market attack into: - target companies - enriched contacts - ranked accounts - rejection logic - attack cohorts - messaging angles - campaign-ready exports No bloated stack required.

Open source won developer infrastructure. Agent harnesses may be different. Generalist open source harnesses are great for experiments. But commercial agents need to be accessible for the average person. ... and commercial agent harnesses are starting to win big time.

The next SaaS pricing model is becoming obvious: seats + agent consumption. Seats still matter for identity, access, and permissions. But agents turn one user into hundreds of actions. The companies that make their APIs headless enough to meter that work are going to print.