Eighty six percent of solopreneurs now try AI before hiring, training, or outsourcing. This episode of the Web Guy Nick Podcast looks at what that means for a one person shop: a faster reply, a clean proposal, and a website that answers questions at nine at night. Alex, Jordan, and Nick also cover Gartner’s warning about a 3.5 trillion dollar gap between AI spending and AI value, and Wall Street tying executive pay to AI skills. The rule through all of it: AI first for the draft, human last for the decision. Listen on Spotify, and visit webguynick.com to put that order to work.
Listen to this episode on Spotify
Trying AI first is smart. Trusting AI first is not. If the task touches money, a contract, or a customer promise, a human still checks it before it goes out. AI first for the draft, human last for the decision. That order matters.
Nick
That sixty four percent is the real story. A one person shop does not mainly need to shave twenty minutes off email. It needs to look like a real operation. A fast reply, a clean proposal, a professional follow up, a website that answers questions at nine at night. That is what wins the job against a bigger competitor.
Nick
I tell clients this all the time. Rent the tools, own the data. The model can change next month. Your customer list and how your business actually runs should never be trapped inside one app.
Nick
Full transcript
Alex: Here is the number that should stop every business owner mid-scroll. The AI industry is on track to spend nearly one trillion dollars on infrastructure this year alone, and one of the most respected research firms in tech just said the revenue to pay for all of it may not be there. Not even close.
Jordan: That is Gartner, and we will get to their three point five trillion dollar warning in a moment. I am Jordan, this is Alex, and joining us as always is Nick, who actually runs a business with this stuff every day.
Nick: And that is exactly why today matters. The big companies are spending like the future is guaranteed. Small owners like us have to make every tool earn its keep this month, not someday. We have five stories today, and every one of them comes back to that one question. Is this AI actually paying for itself?
Alex: Story one is the Gartner report, published today, Wednesday, October fourteenth. Gartner says AI providers are expected to spend nearly one trillion dollars on infrastructure in twenty twenty-six, and another six trillion dollars between twenty twenty-seven and twenty thirty.
Jordan: And here is the other side of the ledger. Gartner says data center owners would need to generate eight point five trillion dollars in cumulative revenue through twenty thirty just to cover the spending, the financing costs, and a twelve percent return on the capital invested.
Alex: But the forecast for enterprise and consumer AI revenue that can actually be tied to those data centers is no more than three point six trillion dollars over the same period. That is a gap of almost five trillion dollars between what gets spent and what comes back.
Nick: Let me put that in shop terms, because trillions stop meaning anything. That is like a contractor buying ten trucks, a warehouse, and a full crew on the bet that the phone will ring enough to cover all of it. Maybe it will. But if the calls come in slower than the payments, somebody writes down a lot of expensive equipment.
Jordan: Gartner is blunt about who gets hurt first. By twenty thirty, they predict more than eighty percent of the smaller cloud providers, the ones called neoclouds, will cease operating independently. Cost pressure and no clear difference from the big players squeeze them out.
Alex: Which lands on your business in a practical way. If your AI tool runs on a small provider that gets bought or shut down, your workflow can vanish with it. So the tip here is simple. Keep your important prompts, your customer data, and your files exportable. If you could not move to another tool in one afternoon, you are more locked in than you think.
Nick: I tell clients this all the time. Rent the tools, own the data. The model can change next month. Your customer list and how your business actually runs should never be trapped inside one app.
Jordan: Story two is closer to home, and honestly it might be the most important hiring story of the year for very small businesses.
Alex: FreshBooks published a new report this week called The Era of the Solopreneur, based on a Wakefield Research survey of five hundred solopreneurs and microbusiness owners. The headline number is eighty six percent. When a task comes up that they cannot finish entirely on their own, eighty six percent say they try AI before they hire someone to do it.
Jordan: And it gets more interesting. Asked which capability would matter more to their future success, sixty four percent picked the ability to appear bigger and more capable than their actual size. Only thirty six percent picked simply getting more done without adding time, employees, or costs.
Nick: That sixty four percent is the real story. A one person shop does not mainly need to shave twenty minutes off email. It needs to look like a real operation. A fast reply, a clean proposal, a professional follow up, a website that answers questions at nine at night. That is what wins the job against a bigger competitor.
Jordan: A concrete example. You are a solo realtor and a listing inquiry comes in while you are at a showing. An AI assistant drafts the reply, pulls the showing times, and books the tour. You review it in thirty seconds between appointments. To the client, you look like you have a full office behind you, because in a sense, you do.
Nick: But here is the owner’s warning that goes with it. Trying AI first is smart. Trusting AI first is not. If the task touches money, a contract, or a customer promise, a human still checks it before it goes out. AI first for the draft, human last for the decision. That order matters.
Alex: Story three comes from Wall Street, where AI skills just became a pay issue. Reuters reported today on a survey by Selby Jennings, a financial services recruiting firm, which polled five hundred executives across the finance industry.
Jordan: The findings are striking. Nearly a quarter of financial firms now tie raises, bonuses, or promotions to how proficient employees are with AI. Another thirty percent require employees to disclose their use of AI tools in formal performance reviews.
Alex: And forty two percent of respondents said their firm has introduced formal criteria to evaluate employees’ AI skills, while thirty percent have created dedicated teams or roles to track whether their AI investments are actually delivering results.
Nick: People hear Wall Street and think this does not apply to a ten person company. It does, just in a smaller way. You do not need a formal bonus formula. But if one employee can produce a week of social posts, clean a messy spreadsheet, and draft customer replies in a day because they learned the tools, that skill is worth recognizing, and worth teaching to everyone else.
Alex: The takeaway for smaller teams is clean. Pick one measurable thing per person. Faster quote turnaround, fewer errors in invoices, more reviews answered the same day. Then check in a month whether the number moved. That is the Wall Street idea, scaled down to a business you can actually run.
Nick: Story four is for anyone who lives inside QuickBooks, which is most of the owners listening. Intuit announced today that Intuit Intelligence is now rolling out internationally, across QuickBooks and the Intuit Accountant Suite, with the new features in beta for QuickBooks Online and QuickBooks Online Advanced customers in those markets.
Jordan: Three features stand out. Automations lets you describe a recurring task in plain conversational language instead of building it rule by rule. Autonomous Invoicing creates and sends invoices from customer documents and manages the reminders. And Conversational Custom Reports lets you ask for a report in plain words and see projected cash flow built from your ledger history, bank activity, and predictive models.
Alex: Intuit says the platform draws on data from ten million customers and tracks about six hundred twenty five thousand attributes per business on average. One feature owners will care about most is a cash shortfall flagged before it happens, not discovered after the account is already short.
Jordan: Intuit also shared early usage numbers. Adopters reported saving an average of thirteen hours a month, and seventy percent said they were confident their books were done correctly.
Nick: Thirteen hours a month is real money if it is true for your shop, and the cash flow warning is the piece I would test first. Every owner has had that surprise week where payroll, rent, and a tax payment all land at once. A tool that taps you on the shoulder two weeks early is worth more than a tool that writes a prettier invoice. But test it against your own numbers for one month before you rely on it. A forecast you have not checked is just a guess with better graphics.
Alex: Story five is about who runs the show when you hand work to an AI agent. New survey data from VentureBeat Intelligence looked at agent orchestration, which is the software layer that hands tasks to agents, calls the tools they use, and keeps every agent inside its permissions.
Jordan: Among companies that use an agent orchestration platform, forty three percent said they use OpenAI’s, twenty two percent said Anthropic’s, and twenty one percent said Google’s. So OpenAI holds roughly a two to one lead at the platform layer.
Alex: But the useful story sits underneath. The same research found that changing the model inside a live application costs far more than most teams budget for.
Nick: Translation for a small owner. Do not build your whole business on one agent and assume you can swap the brain in a weekend. You cannot. Choose based on where your work already lives. If your email, files, and calendar are in Google, start there. If your team already pays for one platform, pilot there first. And keep the permissions tight, because the orchestration layer decides what an agent may touch.
Jordan: Time for Higgsfield Corner, and today video just got a lot longer.
Alex: Higgsfield said this week that it can now take AI generated video from roughly one minute of output to more than ten minutes, using new OpenAI models including GPT six Astra for longer generation. Two companion pieces matter too. GPT Image two point five is aimed at more precise visuals, and GPT Live points toward voice driven filmmaking, where you shape scenes by talking instead of clicking through menus.
Nick: Here is why that matters for a local business. A one minute clip is a social post. A ten minute video can be a full service walkthrough, a training video for a new hire, or a proper explainer on your website that answers the questions customers ask every week. That is a different kind of asset, the kind that keeps working for a year.
Jordan: The caution is the same as always. Longer video means more chances for a wrong detail to slip in, a hand that looks off, a product shot that does not match the real thing. Generate the long piece, then watch all of it before it goes public. Ten minutes of unchecked video is ten minutes of unchecked claims about your business.
Alex: Now it is time for AI or BS, our quick fire game. Jordan reads the claim, Nick calls it, and we check him.
Jordan: Claim one. By twenty thirty, Gartner predicts more than eighty percent of neocloud providers will stop operating independently.
Nick: That one sounds dramatic, but I have been listening, so I am calling it AI, it is true.
Alex: Correct. That is straight from today’s Gartner release.
Jordan: Claim two. In the FreshBooks survey, most solopreneurs said they hire help first and only try AI if the hire does not work out.
Nick: No, that is backwards. It is BS. Eighty six percent try AI first. That was the whole headline.
Alex: Correct again. Claim three is mine. Wall Street firms have now banned employees from mentioning AI in performance reviews.
Jordan: I will take this one. That is BS. It is the opposite. Thirty percent require employees to disclose AI use in formal reviews, and nearly a quarter tie pay or promotion to AI skill.
Nick: Two for two for the hosts today. The game is rigged and I am fine with it.
Alex: Which brings us to One Thing to Try This Week, and today it is the AI before hiring test, done safely.
Jordan: Pick one task this week that you were about to hire out, outsource, or just leave undone. Bookkeeping cleanup, a week of social posts, a follow up email sequence, anything with a clear finish line.
Nick: Give the AI a fair, structured shot. Thirty focused minutes. Write down what good looks like first, give it your real examples, and judge the output against that list, not against a feeling. If it gets you eighty percent of the way, a human finishes the last twenty percent and you just saved real money. If it fails, now you know exactly what you are hiring for, and your job post writes itself.
Alex: So the action is simple. One task, thirty minutes, one written scorecard. Try AI first, keep a human last, and measure what you actually saved.
Jordan: That is the show. Gartner says the big spenders need revenue that may not arrive. Solopreneurs are trying AI before hiring. Wall Street is paying for proven AI skill. QuickBooks is flagging cash trouble early. And the agent platforms are settling into place. Spend like an owner, not like a headline.
Nick: And if you want help sorting which of these tools would actually earn their keep in your business, you can find us at webguynick dot com. We will see you tomorrow.
