Why AlixPartners' Co-CEO Is Bullish on AI, and What It Means for Your Business

Talk to most business leaders about artificial intelligence today, and you get one of two reactions. Either AI is about to change everything overnight, or it is an expensive science project that has not paid off yet. Rob Hornby, Co-CEO of the advisory firm AlixPartners, is making the case for the first view, with a twist. His bull case is not built on hype about chatbots or flashy demos. It is built on something far more boring and far more powerful: the slow, unglamorous work of putting AI into the core of how companies run.

That distinction matters more than it might sound. The bull case for AI has shifted. A year or two ago, it was a story about what models could do. Now it is a story about what companies can actually absorb. And the people who advise those companies, the consultants, the turnaround specialists, the operators who get called in when things go wrong, are often the first to see which way the wind is really blowing.

What the Bull Case Actually Says

Strip away the noise, and the bullish argument for AI rests on a few plain ideas.

First, AI is a general-purpose tool, not a product. Electricity did not transform factories because someone sold a better light bulb. It transformed them because every part of the factory was redesigned around the assumption that power was cheap and everywhere. AI follows the same pattern. Its biggest gains will not come from one clever app. They will come from companies rebuilding workflows, job roles, and decision-making around the idea that machine intelligence is now available on demand.

Second, the cost curve is moving in the right direction. Every major wave of computing has gotten cheaper and more accessible over time. When a capability becomes cheap enough, it stops being a specialty and becomes a baseline expectation, like having a website or accepting credit cards. The bull case assumes AI is on that same path.

Third, the adoption gap is the opportunity. Most companies have experimented with AI. Far fewer have rebuilt anything meaningful around it. That gap between experiment and rebuild is where the real money sits. For advisory firms, it is also where the phone starts ringing.

Hornby's position, as a leader of a firm that spends its days inside struggling and transforming businesses, gives that argument extra weight. Consultants see the failures up close. They see which AI projects die, which ones limp along, and which ones quietly change a company's economics. A bull case coming from that vantage point is not a tech-industry pep talk. It is an argument about execution.

Why a Turnaround Firm's View Matters

Advisory and restructuring firms occupy a strange, useful seat in the economy. They get called when a business is under pressure, when costs are too high, margins are shrinking, or a competitor has pulled ahead. That means their view of AI is shaped by a simple question: does this actually fix the problem, or does it just look good in a slide deck?

That lens is valuable because it cuts through a lot of the current confusion. Plenty of companies are buying AI tools and then wondering why nothing changed. The answer is usually that they added a tool on top of an unchanged process. A turnaround mindset forces a harder question: if this technology really works, what should we stop doing, who should we stop hiring, and which steps in our workflow should disappear entirely?

This is the heart of the bull case. AI's value is not additive, it is substitutive. It replaces steps, not just assists them. Companies that treat it as a helper get modest gains. Companies that treat it as a replacement for certain kinds of work get step-change gains. The difference between those two paths is where most of the next few years of business value will be won or lost.

The Shift From Pilots to Production

The clearest signal in the current market is the migration from pilots to production. Early AI experiments were often small, safe, and disconnected from the numbers that matter. They lived in innovation labs and were judged on whether they worked at all, not whether they paid off.

That era is ending. The questions leaders are asking now are harder:

• Which processes are we willing to redesign completely, not just speed up?
• Who owns the outcome when an AI system gets something wrong?
• What does our data actually look like, and can it support the decisions we want to automate?
• How do we measure return when the benefit shows up as avoided work rather than new revenue?

None of these are technology questions. They are management questions. And that is exactly why the bull case is credible. The hard part of AI was never going to be building the model. It was always going to be changing the organization around it.

The data problem is still the bottleneck

Ask any operator why an AI project stalled, and you rarely hear "the model was not smart enough." You hear about messy records, disconnected systems, unclear ownership, and processes that were never documented in the first place. AI exposes the gaps in a company's foundations. It does not paper over them.

This is why the bull case is a multi-year story, not a single-quarter one. The companies that win will be the ones that treat data cleanup, process clarity, and governance as part of the same project, not as a separate IT chore to be handled later.

Workers, Not Just Models

The most underrated part of the bullish outlook is about people. The loudest AI stories are about jobs disappearing. The more useful story is about jobs being reassembled.

In most companies, a single role is a bundle of tasks. Some of those tasks are judgment-heavy and relational. Others are repetitive and rule-based. AI tends to bite hardest into the second group. That does not automatically eliminate the role, but it does change what the role is for. A junior analyst who spent half their week assembling reports may now spend most of it interpreting them. A support agent may shift from answering routine questions to handling only the hard, emotional ones.

That shift is good news for productivity and bad news for anyone who assumed their job would stay the same. The bull case only holds if companies invest as heavily in retraining and role redesign as they do in software. Firms that skip that step will get tools without adoption, and adoption is where the return lives.

Where the Bull Case Could Break

Being bullish is not the same as being blind. There are real ways this story goes sideways.

Trust and accountability. An AI system that makes a recommendation is easy to launch and hard to defend when it is wrong. Regulated industries in particular will move only as fast as their ability to explain and audit decisions. Companies that treat governance as an afterthought will find their rollouts frozen at the worst possible moment.

Measurement failure. If leaders cannot show where the money went, budgets get cut. AI spending that is never tied to a specific business outcome is the most likely casualty of the next round of cost discipline.

Over-automation. Some processes are cheap to automate but expensive to get wrong. Pushing AI into customer-facing or high-stakes decisions before the guardrails exist is a fast way to turn a productivity win into a reputational loss.

Concentration risk. Relying on a small number of outside providers for a capability this central creates dependency. Smart operators will keep the ability to switch, compare, and fall back.

None of these break the bull case. They define the conditions under which it holds. That is a more honest version of optimism than the version that says everything will simply work out.

Actionable Insights: What to Do Now

For business leaders trying to act on this view rather than just nod along, a few practical moves stand out.

• Pick three workflows, not thirty. Choose processes where the volume is high, the rules are clear, and the cost of a mistake is manageable. Go deep instead of wide.
• Define the return before you build. Write down what "better" means in numbers, hours saved, errors avoided, cycle time cut, before any vendor conversation starts.
• Fix the data on the same timeline. Treat data readiness as part of the project, not a prerequisite someone else will handle.
• Redesign the role, not just the tool. For every workflow you automate, name the person whose job changes and plan their transition.
• Build a paper trail for trust. Log what the system did, why, and who approved it. This is cheap now and invaluable later.
• Keep your options open. Avoid locking your core operations into a single provider you cannot easily replace.
• Measure avoided cost, not just new revenue. A lot of AI value shows up as work that never happens. If your metrics only count new income, you will undercount the win.

What This Means for the Future of AI

If the bull case holds, the next phase of AI will look less like a technology story and more like an operations story. The winners will not be the companies with the most impressive demos. They will be the ones that quietly rebuilt their workflows, trained their people, and tied every deployment to a number on the income statement.

That also means the debate will shift. We will spend less time arguing about what AI can do and more time arguing about who is accountable when it does it, how the gains are shared, and what happens to workers whose tasks disappear faster than their roles can be rebuilt. Those are not technical questions, and no model release will answer them.

Hornby's bull case is ultimately a bet on execution, on the idea that the hard, patient, unglamorous work of organizational change will happen. That bet has been right before, in every major technology shift of the past century. It has also arrived later and messier than the optimists expected, every single time. The smart play is to be bullish on the destination while staying clear-eyed about the timeline.

For businesses, the practical takeaway is simple. The era of experimenting with AI is closing. The era of rebuilding around it is starting. The companies that start that rebuild now, with clear goals, clean data, and real attention to their people, will be the ones the next wave of case studies is written about. Everyone else will be the cautionary tale, wondering why the tool that looked so impressive in the demo never changed a thing.