The proposal you spent a weekend on gets summarized by a buyer's assistant before anyone reads it 🐬 Prospects ask a model about you before the first call. And the new SDR you hired is quietly competing with software that never sleeps.

That software gets cheaper every month. Since GPT-3, the price of a unit of machine intelligence has fallen about thirteenfold a year.

By 2030, AI will be close to free. The sales conversation won't be.

So who doubles revenue in that world, and who gets squeezed? And what do you do about it this quarter?

On October 29, 2026, Will Manidis and Joe Roberts published The Commanding Heights, a book-length essay on where the money in the AI era will actually settle. It's written for investors. Railroads, the dollar, Lenin, Guinness. But underneath it sits a map of how buying and selling will work by the end of this decade.

Here are the 5 theses from it that matter most if you grow your company through sales. With what each one means for your Monday.

1. The intelligence is the giveaway. The connection gets paid 🐯

The essay's core claim: new technology almost never makes its inventors rich. Gutenberg went insolvent. The railroad fortunes went to people who "never built a locomotive." The authors expect AI to follow the same pattern. Models get copied within months (open models trail the best closed ones by about 4 months, per the Epoch AI data they cite), and the price of intelligence keeps falling faster than any technology buildout before it, railroads and telephone included.

Their line: "In the world to come, one either owns the rails or pays those who do."

What it means for you: by 2030, your expertise, your playbook, your clever proposal template.. a model will produce something close for almost nothing. What stays scarce is the crossing every deal must pass through: the relationship, the trusted conversation, the moment a buyer decides. That is your rail. Own it.

In plain terms: the people who'd take your call tomorrow, the community that opens your emails, the partners who send you names, and above all the sales conversation itself. If you rent all your attention from platforms and all your process from tools, you're paying the toll 🌴

2. When friction falls, everything comparable gets shopped to the bottom

Banks earn on deposits because customers have better things to do than chase the best savings rate. The authors predict personal AI agents will run a "mini-treasury function" for every household, so deposits "will autonomously glide to their highest return as friction falls."

Now picture your buyer's procurement in 2030. An agent compares five vendors on scope, price and reviews before a human reads a word. Anything legible and comparable gets pushed toward the cheapest offer. You already hear the early version on calls: "Can you just send something over?" Then silence, while your proposal goes stale in a comparison table. The essay says it bluntly about the models themselves: commoditized capability gets priced at "the marginal price of the compute serving it."

What it means for you: pricing power won't come from a better feature list. Agents win that game. It comes from what no agent can compare: the number your buyer names out loud when you ask what doing nothing costs them. That's Step 4 of the 8 Steps of the Repeatable Sale, Cost of Inaction. When a buyer says something like "this costs us $480,000 a year," your price sits next to their number, not next to a competitor's PDF 🐬

That's control of the sales conversation. No begging for the meeting, no chasing the proposal, no discount to win. You lead, the buyer names the stakes, the price makes sense.

3. The company of 2030: a few operators, many agents 🐯

The essay describes four eras of how companies are built: the owner-operator (Arthur Guinness and his brewery), the manager era, the shareholder-disciplined era, and now a fourth. "Tomorrow, firms will be composed of a few operators, paid and governed like capital, directing work done mostly by models."

It's already in the numbers. In fiscal 2026, Tata Consultancy Services cut headcount by 23,460 while adjusted operating profit rose 7.5%. Profit and headcount used to rise together. They're "coming unstuck."

What it means for you: the sales team of 2030 isn't 12 SDRs and a VP. It's 1 to 3 operators who lead conversations and direct agents that research, write first drafts, follow up and log everything. That's good news if you're tired of being the only closer. You don't need a big team to stop being the bottleneck. You need one person who can run the process without you, and a process clear enough to hand over.

Which brings up the uncomfortable part..

4. Whatever can be described gets automated. Judgment and accountability stay human

The essay's sharpest scene: Ravi, a manager in Bangalore, spends months approving his AI agents' output. Then his keycard stops working. "He was in the training environment for the agent that replaced him." Bitter. And worth chewing on, because the same logic runs through your sales team.

The authors' rule, borrowed from economist Ronald Coase: work that is easy to describe and measure moves out of the firm. Research, list building, first-draft emails, meeting notes, CRM updates.. all legible, all heading to models.

Two things stay. First, accountability. "A liable human must remain responsible for the output," they write, and licensing "will retain value for its implied certainty." When everything is machine-written, the buyer pays for a human who stands behind the result. Second, the illegible: reading a pause, sensing who in the room really decides, leading a conversation where the buyer discovers their own problem. In other words, control. An agent can book the call. It can't hold the room.

There's a warning in this for your hiring. A Stanford study they cite found employment of 22 to 25 year olds in AI-exposed occupations 19% below trend by mid-2026, while experienced workers showed no gap. The junior ladder, where SDRs learned sales by grinding calls, is disappearing. Your new hires won't get thousands of practice calls on the job anymore. They'll need deliberate reps somewhere else. Spotters. A gym 🌴

5. The window is now, and it closes fast 🐬

The last thesis is the most urgent. Diffusion has barely started. The authors estimate the people whose work AI has truly remade number in "the low tens of millions." Most of the economy is "still plodding along, unknowing."

But the gap between users is exploding. Among OpenAI's enterprise customers, the top 10% of firms by usage generated 8.3 times as many output tokens per user as the middle 10% in June 2026. In January, it was 2.6 times. Picture the founder in your market who's already in that top 10%. Every month, their lead over everyone else gets wider.

Early speed alone doesn't hold. Their example: Malena, a designer who tripled her client list with AI.. until the model learned from her work and her clients realized they didn't need her anymore. Being early with a tool is a lead of months. Turning that lead into a position you own is what lasts. "For a moment in time, right now, the positions that will command what comes next can still be bought or built."

What it means for you: use the next 12 months to turn AI speed into something durable. More conversations started, more relationships owned, a process your team runs without you. Not just the same work done faster.

What a sale looks like in 2030

Put the five together and you get a Tuesday in 2030:

• Your buyer's agent has compared you before the first call. Features and price are table stakes.
• The deal is decided in the conversation, by whoever leads it. Not by the slickest deck.
• Your team is small: an operator or two, plus agents doing everything legible.
• Your price holds because the buyer named the cost of doing nothing, in their own words.
• Your growth runs on rails you own: relationships, community, referrals. Not rented attention.

The founders who win won't be the ones with the most AI. AI will be close to free. They'll be the ones in control of the conversation AI can't have for them 🐯

That's what we practice with founders in the Sales Gym: the 8 Steps of the Repeatable Sale, in order, with spotters who push you through your current sales wall. More than 2,400 B2B founders have gone through our programs. The 8 Steps get more valuable as AI gets cheaper, because they live in the one place AI can't go for you.

Sprintlife is exactly that: control in sales plus time freedom. Agents take the grind. You keep the conversations that matter, and your evenings.

Your one move this week

Take your last 3 deals. For each one, write down whether the buyer said, out loud, what doing nothing would cost them. If they didn't, that's the conversation to practice before 2030 arrives early 🐬 Then do it again next Friday. Three deals, one honest line each. It becomes the scorecard you look forward to.

Want the tools, the AI skills and a community of founders already building this? Start your 7-day free trial of Sprint Club: https://www.strategysprints.com

Ready to accelerate now? Book a Discovery Call: calendly.com/strategysprint/discovery-call

Happy hunting. Simon & The Sprinters 🐬⚡️🐆

Source: Will Manidis and Joe Roberts, The Commanding Heights, October 29, 2026, commanding-heights.com