Zac Kass, former OpenAI top executive: AI will create many market winners

Zack Kass, former go-to-market leader at OpenAI, believes that the artificial intelligence market is still in an early phase, which means the main winners are just beginning to emerge. At the Citi Macro Conference in Hong Kong, he sent the market a simple signal: the time to watch from the sidelines is ending.

The market is not yet divided

Kass views AI not as a story about one undisputed champion that will take the entire market, but as a long cycle of capability redistribution. If the technology is truly in its early stages, then major wins have not yet been locked in, and leadership will develop not only at the level of the models themselves, but also at the level of applications, services, infrastructure, and business processes. This is an important signal for the market: the window of entry has not yet closed, and the decision to "wait until everything settles" could turn out to be more risky than a careful launch right now.

"AI will create many winners."

This thought came from someone who spent a long time commercializing AI technologies and saw how companies make decisions not in theory, but in real operational environments. So the emphasis here is not on a catchy slogan, but on applied logic. While some argue whether the sector is overvalued, others are learning to embed models into everyday work, testing scenarios, and capturing the first effects. At this stage, winners are more often those who turn technology into measurable results faster than those who shout the loudest about strategy.

There will be more than one winner

What's important in Kass's words is the structure itself: AI, in his view, won't necessarily create one new monopolist, but it is capable of forming a broad group of strong players. This resembles previous technological shifts, where not only the developers of the base platform won, but also those who packaged it better into a convenient product, embedded it in industry context, or radically improved service economics. In other words, value will accumulate across the entire chain, not in one narrow layer.

For business, this means several obvious zones of growth:

• launching new AI features in existing products
• accelerating sales and marketing through automation
• reducing the burden on support and back office
• creating industry-specific solutions with clear ROI
• rapidly testing hypotheses without long development cycles

From this logic follows a more pragmatic conclusion: a winner can be not only someone who builds their own model, but also someone who knows how to properly apply someone else's. For many companies, this is even a more realistic path. Not everyone needs their own fundamental AI stack, but almost every business would benefit from understanding where a model improves conversion, reduces operation time, increases customer retention, or opens a new service format. This is where most of the commercial value will be played out.

What business should do now

The main practical conclusion from Kass's position is simple: the period of observation is ending. If the market really is early, then companies still have a chance to occupy a strong position without exorbitant budgets, but to do this they need to move through concrete cases, not through abstract "AI strategy" on slides.

The usual first step is to choose an area with a clear metric where you can quickly see the effect: speed of response to a client, cost of a lead, time to prepare a document, accuracy of request classification, or team productivity. The next step is execution discipline. Winners in the AI cycle are likely to differ not only in their access to models, but also in the quality of implementation: data, processes, interfaces, integration into the work flow, and a culture of experimentation.

It's no longer enough for business to simply buy a subscription to a trendy tool and announce that transformation has begun. You need area owners, short iterations, clear KPIs, and a willingness to cut what doesn't produce results. An early market is good because it gives an edge to those who can learn quickly.

What this means

Zack Kass's thesis is useful because it reduces the heat of a false choice between "everything is already divided" and "we urgently need to build our own model." Most likely, the main battle in the coming years will be over the quality of implementation. For business, this is good news: there are many forms of victory in AI, and some of them are accessible right now.