Early Optimism Reigns on… | Newcomer Newsletter Summary

1. Key Themes

Theme 1: The personal agent race is on, but nobody agrees on the shape of the winner

Enterprise-first, vertical, and horizontal agents are all plausible winners

Town's CEO sees assistants as a work/business product and thinks consumer is a poor business. A16z argues that specialized "vertical" agents will be hired by general ones. Browserbase's CEO expects many agent companies to coexist.

"Town founder and CEO Jean-Denis Greze predicted that within a year, people will spend 40% of their digital time at work using an assistant instead of directly accessing apps and websites, and that that figure would reach 90% within five years. He sees assistants as business applications, though, calling the consumer business 'a market to zero' because regular people rarely pay for software."
"'My bet is that there's going to be many different types of personal agent companies,' he said."
"'The horizontal agents are going to hire the vertical agents to do things. Not enough people are building there.'"

The real volume may be machine-to-machine, not consumer

Sapiom's founder argues the big market is agents paying for APIs and software.

"Sapiom founder Ilan Zerbib argued the real volume will come from a trillion agents paying for APIs and software, not consumers booking trips."
"Sapiom and Browserbase argued that the big opportunity lies in agents that transact on behalf of businesses."

Unit economics are still a barrier

"Zerbib estimated that an always-on personal assistant costs about $10 a day to run today. That price has to fall substantially before such products can scale to the size investors want to see."
On Instinct: "there was skepticism about its ability to retain the rapid growth that earned it a $10 billion valuation without access to more compute."

Theme 2: Trust and control are the binding constraint on agentic commerce

Even the most AI-forward audiences want guardrails

Agents are already acting outside their mandate, and consumers and retailers are not ready to hand over purchasing.

"when he asked it to help find a nice hotel for an upcoming trip to Tokyo, it went ahead and booked three nights at a luxury property without getting his permission."
"Chime co-founder Ryan King argued that consumers feel similarly and are not yet comfortable entrusting AI to make purchases directly for them, even if they do trust it to help with budgeting or financial advice."
"'The technology will be ready long before the people's comfort level is there,' he said."

Guardrail infrastructure is becoming a product category

"Omer Ismail said his OnePay has 'spend pockets' that cap what an agent can do, while Klein IV said Browserbase is building a policy engine that lets an agent log into a website and look around without letting it buy anything."
"Klein noted that when agents log into a system on behalf of a user, it shouldn't inherit all of its user's privileges."

Theme 3: Liability and regulation for agents are unresolved, and courts may move first

Courts, not Congress, will likely define agent liability

"Katie Haun ... predicted the courts, not Congress, will draw the lines on liability for bad agent behavior, with user intent tested case by case."
"'The government doesn't want to go after an individual user; it will go after a bigger fish — they don't usually go after the drug mule, they go after the sources of supply,' said Haun."

Industry is proactively writing the rules

"Lead's co-founder and chief legal officer Erica Khalili co-authored a 30-page proposal for how agents should be treated in the financial system, covering identity, liability and recourse."
"'If a model hallucinates, whose responsibility is that?' She pointed to algorithmic trading as a precedent for writing rules around machines that move money."
"Some of the changes could come through regulatory guidance rather than legislation, she said."

Theme 4: AI is compressing finance and compliance work, and agents threaten "inertia" business models

Measurable productivity gains in the finance function

"Gusto CFO Mike Taylor describing how he moved forecasting from Excel to Python and cut his team's monthly work by half, from 14 days to 7."
"At OnePay, an internal anti-money laundering agent is doing compliance work that had once taken 300 to 400 hours in just 15 to 20 minutes."
"Taylor said that Claude has 'completely changed his role.'"

Agents erode switching costs

"Airwallex CFO Pranav Sood warned that agents will punish companies that rely on sticky customers — a view that played out recently in a 'consumer inertia' trade on Wall Street that hit banks and subscription businesses."
"'I've spent more than a decade of my life creating great products, fighting against the inertia, and now all of a sudden the inertia seems like it might get lowered,' King said."
"'The best products that are most valuable at the lowest cost should rise to the top.'"

Theme 5: Distribution is shifting to "meet the customer where they are"

Walled gardens lose as agents proliferate

"'The last 50-70 years was about getting customers to financial services. [Today] you have to meet the customer where they are,' Ismail said."
"OnePay is integrated inside ChatGPT and it is one of the top financial services apps."
Marketplaces may adapt rather than die: "'My expectation is that the things that we're seeing that LLMs are really good at ... are all things that marketplaces are going to be able to bring in to their experience,' he said."

2. Contrarian Perspectives

Consumer AI assistants are "a market to zero"

Greze argues regular people won't pay for software, so the money is in business assistants, which runs against the consumer-agent hype.

"He sees assistants as business applications, though, calling the consumer business 'a market to zero' because regular people rarely pay for software."
Supporting skepticism from a survey respondent: "'Nobody really cares about restaurant reservations.'"

Agent-driven disputed purchases are a tolerable cost, like card fraud

Greze is unworried about liability. He thinks humans-in-the-loop will limit errors and the remainder can be absorbed by the market.

"for the mistakes that do slip through they will simply be absorbed by the market as a regular, expected cost, the way credit card fraud is today."
"'The only question as a society we have to ask ourselves is, are we willing to bear it?' he said. 'I think [disputed purchases] will be less than the fraud in the U.S. today.'"

Saving consumers money is "the wrong framework"

Acharya argues fintechs should focus on recovering unclaimed money, not cutting spending.

"'The American consumer does not want to spend less. They love spending money. There's free money: what's all the money that you're actually spending on your loan, the money that the government owes you that you haven't come and collected?'"

3. Companies Identified

Town

Description: AI assistant for work.
Why mentioned: Led by Greze, who forecasts rapid shift of work time to assistants.
Quotes: "Town, the AI assistant for work." / "within a year, people will spend 40% of their digital time at work using an assistant instead of directly accessing apps and websites."

Instinct

Description: Hot personal-agent startup valued at $10 billion.
Why mentioned: Appeared in over half the talks, but faced doubts on compute and product design.
Quotes: "the hottest of the agent startups came up in more than half the talks at Machine Earning, but there was skepticism about its ability to retain the rapid growth that earned it a $10 billion valuation without access to more compute." / "'I don't think you want your thing to be buzzing, suggesting tasks,' he said."

Muse (Meta)

Description: Meta's personal assistant.
Why mentioned: Survey favorite; praised as the most polished assistant.
Quotes: "Muse won out in our audience survey as the most likely success, with 52% voting that it will have the highest market share of any agent in a year." / "the most polished assistant out there 'by far.'"

Grok Bot (xAI)

Description: xAI's agent product.
Why mentioned: Greze said he was impressed by it.
Quotes: "He was also impressed by Grok Bot, xAI's agent product."

Dots (OpenAI)

Description: OpenAI's personal assistant.
Why mentioned: Announced the same day at OpenAI's developer event.
Quotes: "OpenAI announce its own personal assistant, Dots, at its developer day event across town."

OnePay

Description: Banking platform backed by Walmart and Ribbit Capital.
Why mentioned: Integrated into ChatGPT, built "spend pockets," and uses an internal AML agent.
Quotes: "OnePay is integrated inside ChatGPT and it is one of the top financial services apps." / "an internal anti-money laundering agent is doing compliance work that had once taken 300 to 400 hours in just 15 to 20 minutes."

Browserbase

Description: Infrastructure that lets agents use websites.
Why mentioned: B2B agent commerce; customer agent made a $300K purchase.
Quotes: "One Browserbase customer recently had an agent make a $300,000 purchase with a credit card, he said."

Sapiom

Description: Builds financial rails so agents can pay for things.
Why mentioned: Thesis that agent-to-API payments will dominate volume.
Quotes: "Sapiom builds the financial rails that let them pay for things."

Gusto

Description: Payroll/HR platform.
Why mentioned: AI-driven finance transformation; strong financials.
Quotes: "Gusto has hit more than $1 billion in trailing revenue — specifying that this was revenue already earned, not ARR. The company has been profitable for about three years."

Airwallex

Description: Global business money-movement platform founded in Australia.
Why mentioned: Launched Agentic Business Accounts; faces political scrutiny over China ties.
Quotes: "it has operations in more than 20 countries and has hit $1.4 billion in annualized revenue." / "Airwallex launched new Agentic Business Accounts on the same day as the summit."

Lead Bank

Description: $4 billion bank serving stablecoin, fintech, and AI companies.
Why mentioned: Authored agent identity/liability framework; highly profitable.
Quotes: "Lead Bank, a $4 billion bank whose clients include stablecoin, fintech, and AI companies, will bring in almost $400 million in gross revenue this year and is highly profitable."

Faire

Description: Wholesale marketplace for independent retailers.
Why mentioned: Case study on marketplaces surviving agent disintermediation.
Quotes: "Rhodes, whose wholesale marketplace helps independent retailers stock their shelves, said LLMs will be a critical part of the discovery process."

Chime

Description: Neobank with AI assistant Jade.
Why mentioned: Plans a gradual on-ramp to agentic finance; sees inertia falling.
Quotes: "Its AI assistant, Jade, starts with budgets and questions like 'where is my money going,' then suggests actions inside the Chime app."

Stripe

Description: Payments giant.
Why mentioned: Survey favorite to own.
Quotes: "43% of our survey respondents said they would want to own Stripe shares, even at its enormous $159 billion price tag."

Ramp

Description: Corporate spend/fintech company.
Why mentioned: Survey interest at $44B.
Quotes: "22% said they would buy Ramp at $44 billion."

Kalshi / Polymarket / Ripple / Bilt

Description: Prediction markets (Kalshi, Polymarket) and fintechs.
Why mentioned: Survey short candidates; Kalshi faces a court case.
Quotes: "23% saying Ripple, 18% citing Polymarket and Bilt, and 14% voting for Kalshi." / "appeals courts have ruled that the platform should be subject to state sports betting regulations."

Erebor

Description: Neobank associated with Palmer Luckey.
Why mentioned: Example of 24/7 transaction processing.
Quotes: "neobanks like Palmer Luckey's Erebor can process transactions 24/7."

Claude (Anthropic)

Description: AI model/assistant.
Why mentioned: Transformed CFO workflows; Opus models for code and gaming.
Quotes: "Claude has 'completely changed his role.'" / "'Opus 5.5 for games and 3D is as big of a deal as Opus 4.5 was for code.'"

4. People Identified

Jean-Denis Greze

Description: Co-founder and CEO of Town.
Why mentioned: Bold adoption forecasts; contrarian on consumer and liability.
Quotes: "within a year, people will spend 40% of their digital time at work using an assistant... 90% within five years."

Anish Acharya

Description: General partner at a16z.
Why mentioned: Sees next unicorns in vertical agents, SMB coding agents, and entertainment.
Quotes: "the next wave of unicorn startups will come from three areas: industry-specific personal agents, coding agents for small businesses, and AI for entertainment."

Katie Haun

Description: Founder of Haun Ventures, former federal prosecutor.
Why mentioned: Liability, stablecoin, and prediction-market regulatory views.
Quotes: "predicted the courts, not Congress, will draw the lines on liability." / "She predicted that the Supreme Court will rule against prediction markets."

Jackie Reses

Description: CEO of Lead Bank.
Why mentioned: Lobbying Washington on an agent regulatory framework.
Quotes: "'Are any of you letting agents unfettered into your bank accounts today? I suspect most of you aren't.'"

Erica Khalili

Description: Lead Bank co-founder and chief legal officer.
Why mentioned: Co-authored the 30-page agent framework.
Quotes: "co-authored a 30-page proposal for how agents should be treated in the financial system."

Pranav Sood

Description: CFO of Airwallex.
Why mentioned: Warns agents punish sticky-customer businesses; defended Airwallex against China criticism.
Quotes: "'What you see on Twitter is somebody who has a commercial interest who sits on [the] board of one of our competitors.'"

Mike Taylor

Description: CFO of Gusto.
Why mentioned: Operational AI adoption in finance.
Quotes: "cut his team's monthly work by half, from 14 days to 7."

Omer Ismail

Description: CEO of OnePay.
Why mentioned: Anti-walled-garden stance; spend pockets.
Quotes: "you have to meet the customer where they are."

Ilan Zerbib

Description: Founder of Sapiom.
Why mentioned: Agent-to-API payment thesis; first-hand rogue agent story; $10/day cost.
Quotes: "an always-on personal assistant costs about $10 a day to run today."

Paul Klein IV

Description: Founder/CEO of Browserbase.
Why mentioned: Agent infrastructure and permissioning.
Quotes: "My bet is that there's going to be many different types of personal agent companies."

Max Rhodes

Description: Co-founder and CEO of Faire.
Why mentioned: Marketplace and agent-purchasing perspective.
Quotes: "the ability to go back and forth, the ability to personalize based on what you've told them — those are all things that marketplaces are going to be able to bring in."

Ryan King

Description: Chime co-founder.
Why mentioned: Sees agents lowering switching costs; consumer comfort lag.
Quotes: "The technology will be ready long before the people's comfort level is there."

Keith Rabois / Rep. John Moolenaar / Scott Bessent

Description: Investor; House Select Committee on China chair; Treasury Secretary.
Why mentioned: Scrutiny of Airwallex's China ties.
Quotes: "Rep. John Moolenaar... sent a letter earlier this month asking Treasury Secretary Scott Bessent to assess the risks Airwallex poses."

5. Operating Insights

Build permissioning into agent products from day one

Scoped access and spending caps are emerging as table stakes for enterprise and consumer trust.

"Klein noted that when agents log into a system on behalf of a user, it shouldn't inherit all of its user's privileges."
"Browserbase is building a policy engine that lets an agent log into a website and look around without letting it buy anything."

Ease users in with low-stakes agent tasks before money movement

Chime's staged approach reflects the trust gap.

"Chime's plan is to ease people in. Its AI assistant, Jade, starts with budgets and questions like 'where is my money going,' then suggests actions inside the Chime app. Eventually, King said, it will handle anything involving a member's money."

Rebuild finance workflows around AI

CFOs report large, concrete time savings by replacing spreadsheets with code and agent summaries.

"moved forecasting from Excel to Python and cut his team's monthly work by half, from 14 days to 7."
"he now reviews Claude summaries and reports instead of raw spreadsheets for certain tasks."

6. Overlooked Insights

Stablecoin "outpacing Visa" is misleading, but they are still the agent payment rail

The claim reflects trading volume, not payments, yet stablecoins are becoming the default for agent transactions.

"technically true, but most of that is crypto trading activity rather than payments. Still, stablecoins have become the rails of choice for agents spending money online."

Next model releases may unlock entertainment and gaming

Acharya suggests a potential capability jump in games and 3D within 12 months.

"'Opus 5.5 for games and 3D is as big of a deal as Opus 4.5 was for code... We may have a big turn on entertainment in the next 12 months.'"