What it takes to become an AI company: Four lessons from the top
The New Stack Walid Abu-Hadba ● Covered by 2 sources
Opinion — commentary, not a factual news event.
Precisely says becoming an AI company is mostly about changing work, not buying tools. The real test is business results, not AI usage stats.
Based on reporting by The New Stack, Walid Abu-Hadba — read the original for the full story.
Summary, retelling and take written by AI under human oversight; images are AI-generated illustrations. How we work · Report an error
Everybody wants to be an AI company now. The messier truth, at least from Precisely’s own shift, is that the hard part is not the models or the tools. It is deciding what the business is actually trying to improve, and then ripping out the old habits that get in the way.
That starts with a business objective, not a pile of software. Growth, faster innovation, better customer relationships, improved margins: those are the kinds of outcomes that give AI a job to do. Without that kind of focus, the technology can just add more tools, more agents, and more activity without making the business any better. AI, in this framing, is not the strategy. It is how the strategy gets executed.
The next trap is acceleration for its own sake. If a process is broken or clogged with approvals, handoffs, meetings, and decision points, AI will only help it go wrong faster. Precisely’s own example is blunt: if an engineer can generate code four times faster, but the code still sits in the same queues, the business does not get four times the value. First remove what no longer matters. Then simplify what does. Only after that should AI be used to speed it up.
Then comes the people problem, which is really the whole problem. Technology alone does not transform an organization. At Precisely, an intern program showed how quickly people’s confidence and working style changed once they could use AI on real tasks. They moved faster, dug deeper, and found out they could do more than they thought. That matters because AI shifts the baseline for everyone, not just new hires. It gives people more room for judgment, relationships, and harder work — but only if the work itself is redesigned around that idea.
The final lesson is about measurement, and this one is overdue. Counting AI users, agents, or tokens tells you usage. It does not tell you value. The article’s argument is refreshingly plain: treat AI more like email or spreadsheets — infrastructure — and measure what the business gets out of it. More capacity, faster delivery, better customer experience, stronger performance. If the main goal is cutting headcount, the company is missing the bigger prize.
My take — AI-written commentary, not fact-checked reporting
This is the sane AI memo every board keeps pretending it already wrote. If the plan starts with token counts and ends with layoffs, it’s just cost-cutting wearing a novelty tie. The smarter move is harder: make the work better, then let the machines do the boring bits.
Read more about this at: The New Stack
