Everything good in life requires hard work, and technology is no exception. Every investment in technology is an investment of time and money made to save time and money.
Each such investment involves three parties: the vendor, the IT staff, and the business owner. The vendor gets the sale, the owner gets the return, and the IT staff earns an income and acquires expertise it can carry to the next job. Normally everyone benefits.
But in IT, as in other complex fields such as healthcare, a shroud of mystery hides the details,* and the gains are not always shared the way everyone first assumed.
That mystery is sometimes deliberate. Complexity can protect an advantage, lure a buyer, or build dependency. A vendor can sell features that do not yet exist, betting that if a client ever tries to use them, they can be built and shipped in an update before the gap is noticed. Its bolder cousin is to claim powers that were never there at all, and that version now has a name and a regulator behind it. In 2024, two advisory firms paid a combined 400,000 dollars to settle the U.S. Securities and Exchange Commission's first cases against "AI washing," both having advertised artificial-intelligence capabilities they did not actually have, one of them billing itself as the "first regulated AI financial advisor."1 The Federal Trade Commission opened a broader sweep of its own months later.
Other tactics are cruder. Beta testing is pushed onto paying clients. An aging codebase is carved into new products under new names to ride a fad or force a wholesale upgrade. The largest players bend standards while paying lip service to interoperability. And sometimes a product is simply retired to make you buy the next one.
The clearest recent example needed no new feature at all, just a calendar. When Microsoft ended support for Windows 10 in October 2025, it left millions of working computers with a hard choice. Windows 11 requires a security chip called TPM 2.0 that many older machines lack, so the upgrade often means buying new hardware; the alternative is to pay Microsoft to keep patching the old one. For a business, those patches cost 61 dollars per device in the first year and double every year after, to 122 and then 244, for up to three years.2 The machines still worked. The support did not.
As Big Tech's software grows, it embeds itself deeper in your infrastructure, until leaving becomes a project in itself. None of this would last in a market where a product could be returned and a vendor dropped or sued.
IT rarely offers that kind of market discipline. There are no ambulance chasers in a field where the same person is often judge and party: the one who recommends the system, builds it, and then bills to keep it running. Their interests and yours can quietly part. A high-maintenance system is job security for whoever tends it, while a simpler, sturdier one can put that person out of work, so a better system is not always in the interest of the person you are paying to choose it.
Obfuscation through complexity, whether natural, accidental, or deliberate, pushes the true cost of technology under the rug and down the road. When it is your own dime and your own time, you come to hate the days lost to problems that were planted, or that grew from plain carelessness.
The newest kind of technology that uses you does not waste your time so much as quietly draw on you. In November 2025, LinkedIn switched on a setting, on by default, that feeds its members' posts and profiles into the training of its own artificial intelligence. You can turn it off, but only from that point forward; what the model has already learned from you cannot be recalled.3 It is worth asking of any tool before you come to rely on it: who learns from what you put in, and how would you know the switch was on? That is how you learn, sometimes the hard way, the difference between the technology that uses you and the technology you can use.
Steering a client toward the technology they can use, and away from the kind that uses them, is most of what our work comes down to. Our Stewards are paid for engineering judgment, so a leaner, sturdier estate that asks less of us is a result we can afford to want. We favor tools a client can inspect, move to other hardware, and walk away from. And we treat where a client's data lives, and who may read or learn from it, as part of the duty we owe, a choice we make deliberately for every client.
Concierge CIO Partners is a unified Guild of senior technologists providing dedicated, long-term fiduciary IT leadership to midmarket service firms. It offers a strategic alternative to fragile internal IT silos and to Managed Service Providers who advise you on what to buy and are paid on what you buy. With transparent unit pricing and an automated service catalog, the Guild eliminates administrative bloat and ensures every IT dollar spent and decision made directly drives your financial performance. Because the Guild's pay turns on its judgment, its interest and yours point the same way: toward the technology you use, and away from the technology that uses you.
* Language, costs, paperwork, side effects and long-term effects. 1 First SEC "AI washing" enforcement actions, 18 March 2024 (Delphia and Global Predictions, 400,000 dollars in combined penalties): SEC, 18 March 2024. FTC "Operation AI Comply," 25 September 2024: FTC. 2 Windows 10 support ended 14 October 2025: Microsoft. Windows 11 requires TPM 2.0: Microsoft. Commercial Extended Security Updates at 61 dollars per device, doubling each year for a maximum of three years: Microsoft Learn. 3 LinkedIn's "Data for Generative AI Improvement" setting is on by default and opting out stops only future training, effective 3 November 2025: LinkedIn Help.