Gabriel Hardy-Françon
We’ve heard this story before.
A new technology emerges. It gets hyped to the moon. Every investor wants in. Every company launches a lab or an innovation unit. The press calls it a revolution.
And then? It fizzles.
It happened with IoT. It happened with VR, NFC, crypto, and a dozen other "game-changers." And unless we're careful, AI might go the same way.
Not because AI isn't powerful. It is. But because — like IoT — it risks missing a key ingredient: distribution.
The Internet of Things (IoT) refers to the network of physical objects — "things" — embedded with sensors, software, and connectivity to collect and exchange data with other devices and systems over the internet. Think smart thermostats, fitness trackers, connected fridges, industrial sensors, and autonomous vehicles. IoT promised a seamless, data-driven world — but it rarely delivered at scale.
At the TCV Engage Summit, Casey Winters gave one of the most useful distinctions I’ve heard:
A true platform shift needs two things — a technological shift (new ways of solving problems) and a distribution shift (new ways of reaching people).
Let’s break that down:
These pairings enabled massive waves of product and company creation.
Now compare that to IoT:
The result? Some industrial gains, but no breakout consumer adoption.
Crypto is similar. Tokens were a novel form of distribution, but most of the tech didn’t solve real-world problems. It became a distribution engine for hype, not value.
As Kevin Kelly once wrote in The Inevitable, "The future happens gradually, then suddenly" — but only if infrastructure and incentives align. Without widespread usability and adoption mechanisms, a technology stalls at the margins.
Artificial intelligence is clearly a technological revolution. It enables things we couldn't build before. Calk AI wouldn't be possible without it.
But distribution? Not yet.
ChatGPT is not Google. There is no default discovery engine for AI tools. No viral loop like mobile had. Just an overwhelming flood of products, most of which users abandon after one try.
As Benedict Evans notes, tech waves become real when they become boring — embedded in workflows, assumed rather than discussed. AI isn’t there yet.
That means startups can’t rely on the platform to deliver users. They have to build distribution the old-fashioned way:
This is especially true for AI startups, where the hype curve is steep, and attention spans short.
Incumbents (like Microsoft, Notion, and Google) are winning by injecting AI into already-adopted workflows. New players? They need a plan.
As Andrew Chen put it in The Cold Start Problem, "The network is the product." Without user loops and adoption triggers, technology doesn’t scale. AI tools today are still largely single-player, non-networked, and replaceable.
And while many companies are building custom GPTs as quick-fix solutions, these often become siloed, hard-to-maintain interfaces. At Calk, we believe there's a better way: designing flexible, orchestrated agents that work across tools and functions. If you're wondering how to make an AI that lasts, it's not about widgets — it's about workflows.
If you're building in AI, here's the hard truth:
Technology alone won't save you. Distribution is the battlefield.
This is why Calk exists. We’re not betting on a magic new platform to carry us. We're building for immediate value, deep integration, and stack orchestration.
We believe that AI should:
In short: AI shouldn’t feel like a new dashboard. It should feel like a teammate.
That belief is personal for me.
Before joining Calk, I spent years as a product marketing manager. My obsession was always the same: making products feel right to the user. It wasn’t enough for the tech to be powerful — it had to be intuitive, meaningful, and worth coming back to.
Feedback loops? They’re not just a feature to me. They’re the bread and butter of every successful solution. I’ve seen firsthand how user trust is earned — not through features, but through responsive design, thoughtful onboarding, and moments of "this just works."
That’s the mindset we bring to Calk. We don’t want users to build complex agents from scratch. We want the product to greet them like a teammate who already knows the context. If you're asking yourself how to use AI in your business, or how to implement AI in business without reinventing everything, this is our answer: start with the workflows.
Inspired by Crossing the Chasm, we focus on early adopters with real pain points: product marketers, support leads, sales engineers. Not generic productivity seekers, but professionals seeking leverage.
Most of the 2023 AI wave will vanish. Just like IoT dashboards, VR goggles, and blockchain for voting. Not because they weren’t innovative, but because they didn’t combine technology with distribution.
But the winners? They’ll be the ones who stayed focused on:
If you're searching for the best artificial intelligence for your stack, the answer isn't about model size. It's about fit, function, and feedback. That’s how the next generation of artificial intelligence companies will last. That’s what we’re building toward at Calk.
As Jim Barksdale, former CEO of Netscape, once said:
"There are only two ways to make money in business: One is to bundle, the other is to unbundle."
Calk is bundling orchestration, memory, and agents into one outcome-focused interface. Not to chase trends — but to build the bridge between power and practicality.
You can follow the journey or try it for yourself at calk-ai; Or just reach out — we’re always up for a good conversation about AI strategy, agents, or what IoT could have been. Curious about how to use artificial intelligence to support your teams, or even how to create an AI that adapts to your workflow? Let’s talk.
To truly benefit from AI, it must be connected to your tools and data. Discover how integrating context unlocks real results—from productivity to smarter decisions.
Compare Claude 3.7 and GPT-4o for real business use cases. Discover which AI model fits your workflows best—from document summaries to customer emails—and test both on Calk AI.
Discover how agencies can build a scalable AI strategy by connecting internal tools, using multiple models, and deploying custom agents with Calk AI.
