Google's triple threat for Empowerment Tech, and Stripe's AI payments chess move
Plus: Innovation dies in the gap between discovery and execution
Hi everyone, thanks for coming back to Customer Futures.
Each week I unpack the disruptive shifts around Empowerment Tech. Digital wallets, Personal AI and the future of the digital customer relationship.
If you haven’t yet signed up, why not subscribe:
Join us in person - Customer Futures Meetups!
Are you curious about Personal AI, digital ID and digital wallets? Or are you already working on Empowerment Tech?
Join us for casual drinks and a catch-up about all things ET.
LONDON: Thursday 9th October
From 6pm
Main bar, The Hoxton Hotel, 199 High Holborn, London, WC1V 7BD
COPENHAGEN: Wednesday 22nd October
From 6pm
Bifrost House, Sortedam Dossering 55, 2100 Copenhagen
LONDON: Wednesday 10th December
Christmas meetup from 6pm
Main bar, The Hoxton Hotel, 199 High Holborn, London, WC1V 7BD
Looking forward to seeing you there!
Hi folks,
Well, the creative media sector is having a doozy this week. Have you seen the latest OpenAI release of Sora2? It’s properly mad.
But you need to look behind the dazzling content. Because incredibly realistic digital everything is becoming normalised.
Sora2 is wild for creating the most imaginative, breakthrough digital media and content. But we must be so very careful. To understand what it can also now create without us noticing.
Why?
Take wifi on planes. It’s sometimes slow, sometimes too expensive. But by any measure, it’s a technology marvel. Look back 25 years, and we barely had mobile connections on the street.
Yet on a short flight recently, the person sitting next to me was making a massive and embarrassing fuss about how their internet connection wasn’t working. I found it remarkable that they took their connectivity for granted.
Even though we were 40,000 ft in the air, and travelling at 500 mph.
You see, when technology works, it disappears.
For centuries, our living rooms were lit by candles. Then, as electricity was rolled out to the suburbs, people marvelled that they could just flick a switch and light up the house. It changed how we lived, how we shared spaces, even how and when we slept.
Today we take such things for granted. Electricity ‘just works’. And wifi is expected everywhere.
Soon, AI will also disappear into the background of our lives.
I like Ben Evans’ definition of how we should think about it: “AI is whatever doesn’t work yet, because once it works it’s just software”.
Calculators, spreadsheets and spell-check tools were once ‘AI’ marvels, too. And only a couple of years ago, it was an ‘AI’ miracle that you could ask your phone to bring up photos of a specific person from a gallery of 20,000 pics. It’s now just part of the phone OS.
Yes, GenAI is now here. But it’s about to disappear, just like electricity and wifi. And we won’t notice it any more.
So it’s precisely at this moment - when the tech disappears - that we must be most cautious about the risks.
Creative AI tools are going to be embedded everywhere, and into everything. Once it’s ‘just software’, AI soon gets added to every digital product, every digital channel, every online session, and every digital transaction.
AI everywhere.
Today, AI agents are the next breakthrough. The next AI wave. But what happens once they too are normalised? When they too disappear?
We need to be so very careful.
Because our Personal AIs are going to know us better than we know ourselves. Today it’s a monstrously clever AI video, with added audio. Tomorrow, it’s going to be AI Agents that can do things without us noticing - maybe even without us caring.
Personal AI Agents are going to be amazing, helpful, exciting. But they’re also going to be personal and intimate. And have the potential to expose us to massive new levels of manipulation. Of discrimination. And likely many other ethical issues we can’t even yet see.
So, yeah, risks. Many of them.
Some say “don’t worry, we’ll manage those AI risks, we’ll design for them, because we’ll be aware of them.”.
But I’m not so sure. Even when we know there are risks - when we can even see them - our brains won’t be able to deal with them rationally.
Let me give you an example. I love looking back at this video of the ‘hammer experiment’ to remind me how easily our brains can be deceived. Even when we know it’s not our real hand being hit with a hammer.
So as we rush to add Personal AIs to our lives - perhaps for hyper-personal healthcare, or financial planning, even relationship guidance - we need to make sure that these powerful new tools are acting in our interests.
And not just benefiting the corporations that run them.
The hammer experiment shows just how easily our brains can be tricked into feeling pain. My biggest worry is that Personal AIs will in fact be able to do the opposite. They’ll hit us in the digital face with a hammer, but we won’t even feel it.
Like wifi in the sky, our AI tools will disappear into the background.
And we need to be paying attention.
So there’s never been a more important time to help shape the future of being a digital customer. So welcome back to the Customer Futures newsletter.
In this week’s edition:
Stripe’s new Agentic Commerce Protocol - a master chess move?
Innovation dies in the gap between discovery and execution
Google’s triple threat for Empowerment Tech - the biggest upgrade to Chrome in its history
… and much more
Go make a brew, find a cosy spot, and Let’s Go.
Stripe’s new Agentic Commerce Protocol - master chess move?
A huge new announcement from Stripe, partnering with OpenAI on a new ‘Agentic Commerce Protocol’ (ACP). Basically, you can now check out instantly from within ChatGPT.
Why you should care: If a business already processes payments with Stripe (that’s millions and millions of companies), they can turn on payments for agents with only a line of code.
One of Stripe’s superpowers has always been deeply understanding the developer experience. They famously sweated the small stuff in the early days, for example travelling miles to physically sit next to their first customers, to actually watch developers using their early API documentation. To get feedback as fast as possible.
It’s another week, and yet another agentic payment protocol announcement. So what?
Stripe just entered the chat.
What if this time, it’s about distribution?
You can now just switch on ‘Instant Checkout’ in ChatGPT. Stripe has already launched with US Etsy sellers, and according to the press release, they’ll soon add over a million Shopify merchants.
Now that’s distribution.
Stripe has the footprint. They have the payment infrastructure. And now they have the OpenAI partnership to add AI agent payments everywhere.
Cleverly, they’ve made it an open standard. Why? Because it gives them even more scale (bold mine):
“Instant Checkout is powered by the Agentic Commerce Protocol (ACP), a new open standard codeveloped by Stripe and OpenAI that enables programmatic commerce flows between buyers, AI agents, and businesses.
“It provides a blueprint for how businesses can make their checkouts agent-ready so that customers using AI agents, such as ChatGPT, can buy products directly from where they’re discovering them.’
Making this an open protocol means more AI agent traffic, more business integrations, and more payment volume for Stripe.
With OpenAI, with ChatGPT, and with this open standard, Stripe just landed distribution.
And Bryan Georges made the most important point of the week:
“This new distribution channel raises major strategic and operational questions for digital marketers and eCommerce leaders around the diminishing role of their websites in a agentic/conversation-first era”
Bryan is bang on the money.
When any digital touchpoint can now integrate AI Agent payments, the chessboard just changed.
Because if you are a business selling things online (that’s almost all consumer-facing businesses in the world), what happens when your ‘sales funnel’ just disappears? What happens when you can’t show ads to the customer anymore, because their Personal AI ignores them?
Strategic and operational questions indeed.
But there are many more questions coming.
How will we really know that it’s really YOUR agent placing the order with Stripe? Yes, the payment is approved with your details, but you can bet that MalBots will be taking and using your data too, no?
What happens when it goes wrong, and the AI buys stuff you didn’t want? And how much does the AI agent need to know to be effective? Your delivery address? Your shoe size? Your shopping history and preferences? Your availability to be at home?
A few years ago, Gartner predicted that this automated customer market could be “twice the size and arrive at twice the speed of eCommerce.”
Agentic commerce is going to be a big deal. So buckle up folks. We now have agents. We now have payments. And Stripe just unlocked distribution.
Yet another AI chesspiece to track.
The chessboard is getting pretty busy.
STRIPE’S ANNOUNCEMENT, BRYAN’S TAKE
Innovation dies in the gap between discovery and execution
I loved this post about why enterprises almost always don’t catch the disruptive wave that happens right under their nose.
Not because they can’t see it - in fact, many lead the disruption - but because they can’t execute.
Melissa Perri had a front row seat to the real story of why Kodak didn’t win in digital photos. And it wasn’t that they couldn’t see the commercial opportunity. Or, as business legend has it, because they were ‘addicted to film processing fees’.
No, it was about execution.
“Having the right discovery process means nothing if your organization can’t execute on what you discover.
“Kodak had customer research, market validation, and technical capability. They didn’t have organizational structure to act on it.
“Innovation dies in the gap between discovery and execution. While Kodak waited for budget approvals and team formations, Instagram launched and captured the exact opportunity we’d identified.
“The next time your team discovers a validated opportunity, ask yourself: can we move on this now, or are we just another Kodak waiting for the perfect conditions?”
Ideas are cheap folks. Execution is everything.
But I’d add one more ingredient to winning the Next Big Thing. You also need distribution.
Insta had some great viral features, but it really took off because they got bought by Facebook, and rode the social wave.
Ideas + execution + distribution.
You need all three.
I’ve worked with plenty of large organisations that had both the ideas (eg powerful new use cases with customer data, digital wallets and credentials) and the distribution (MM of ‘daily active users’), but they couldn’t execute.
And I’ve worked with and for many companies that had world-class execution and incredibly disruptive and valuable ideas… but they didn’t have distribution.
In the Empowerment Tech market, how many digital wallet and credentials projects have you seen come and go? How many data stores and vaults? And today, how many Personal AI projects will never make it to market?
The Empowerment Tech market is bubbling. We’re not short of ET ideas.
It’s just that we’re not yet there on product execution, which in many ways just means ‘value’. Too many companies are building products when they should be building features for other services.
But most of all, we’re short of distribution.
It’s always about ideas + execution + distribution.
Who’s going to win with all three?
Google’s triple threat for Empowerment Tech - the biggest upgrade to Chrome in its history
Well now. Google has finally built Gemini into the browser.
Chrome with AI can now read your open tabs, answer questions about the web pages you visit, assist with maps, and of course work across Google Docs, Calendar and Slides.
And apparently, AI agents are coming as a fast-follow feature.
With Gemini (or presumably ‘Mariner’, their Agent project?) doing stuff on your behalf. Like handling complex shopping flows, reservations and bookings, and interacting with information in your documents and your diary.
Is this the early view of what mainstream adoption of AI Agents looks like?
Stuart Winter-Tear makes some great points about where this goes next for Chrome:
“Scoped delegation: grant time-boxed, task-scoped permissions and revoke them cleanly after.
Standardised site intents: sites expose high-level actions (book_appointment, pay_invoice) instead of brittle pixel-clicking.
Receipts and rollback: every meaningful step produces a signed receipt and a one-click undo path.
Enterprise guardrails without killing UX: Workspace-grade policies and logs alongside useful defaults.
Human-in-the-loop by default: previews, diffs, and sensible guardrails for high-impact steps.”
And he’s clear-eyed about the implications.
“When an agent can book, buy, or change credentials, it starts to look like root access to your life.
“That’s not a reason to stall - it’s a reason to design for least-privilege delegation, transparent receipts, and fast revocation from day one. Treat page content as adversarial, treat capabilities as time-scoped, make undo a first-class action.
“Chrome just lowered the friction between reading, deciding, and doing. Pair this momentum with scoped delegation, standardised intents, and auditable actions, and we move from “AI that helps” to AI you can trust with real work. For the mainstream.”
Yup. This is actually an excellent checklist for Personal AI Agents and Empowerment Tech generally:
Least-privilege delegation
Transparent receipts
Fast revocation
Treat content as adversarial
‘Undo’ as a first-class action
I’ll add one more observation.
Google also has the Digital Wallet. Which already has support for government ID documents, employee ID, tickets and more. And soon, qualified digital signing.
Bring these things together - the browser (the interface and context), the wallet (trust), and agents (delegation), and you can see quite a formidable combination of services.
Does Google now have the triple threat for Empowerment Tech: the idea, the execution and distribution?
OTHER THINGS
There are far too many interesting and important Customer Futures things to include this week.
So here are some more links to chew on:
Article: Prototypes used to answer “will it work?” but now they need to answer “will people trust it?” READ
Idea: When Customers Set the Terms: How the Intention Economy and ‘MyTerms’ Enable the Great Unwinding READ
Post: The elephant in the room for the EUDI Wallet: monetization READ
Article: What is the difference between a government digital ID, a citizen ID number and a citizen profile? READ
Post: AI means experts get jetpacks and beginners get blindfolds READ
And that’s a wrap. Stay tuned for more Customer Futures soon, both here and over at LinkedIn.
And if you’re not yet signed up, why not subscribe: