The Boardroom Disconnect: Why Your AI Strategy is Stalling at the Top
The conversation around AI transformation has hit a critical bottleneck, and it is not about the technology. It is about the leadership. A major new study published this week by Boston Consulting Group (BCG) reveals a stark and growing divide between CEOs and their boards on how to navigate the AI era.
If you are a founder or C-suite leader, the findings from BCG’s Split Decisions: The BCG CEOs and Boards Survey should be a wake-up call. Based on responses from 625 leaders globally, the data exposes a fundamental misalignment in expectations, accountability, and understanding at the very top of the organisation.
The Era of the Single-Point AI Tool is Over
Google for Startups has just published its 2026 ‘Future of AI’ report, and it is one of the most honest documents about where generative media is actually heading that I have read in a long time. Eight founders and investors, all building or backing companies at the frontier of AI, were asked to share their unfiltered views.
What came back was not a list of features but a set of arguments about what it actually means to build something defensible in a world where the tools are available to everyone.
Here are four ideas I have not stopped thinking about.
1) You Are Not Building Tools Anymore
Sami Ede, Co-Founder of Leonardo.Ai, puts it plainly: “You’re not building tools anymore. Rather, you need to build entire workflows that capture the user end-to-end”.
The era of the single-point AI tool is over. What he is describing is a shift from products that do one thing well to products that own the entire experience. Think about what that means in practice. An agent that transcribes your creative brainstorming meeting, then goes off and executes the ideas from it, generates the content, stitches it together, and hands it back to you for sign-off. The human steps in for curation and final judgement. The agent handles the bulk of the work. That is not a tool. That is a workflow.
2) Taste is the Only Thing AI Cannot Copy
This is the argument that runs through the entire report, and it is worth sitting with. Joaquin Cuenca Abela, Co-Founder of Freepik, makes a distinction that most people miss: “You can’t plug something in and get a perfect copy of someone else’s mind. Humans have this unique, uncopyable element, our past experiences and everything that shaped us into who we are”.
Grace Wang, Co-Founder of OpusClip, frames it in terms of what will actually win on social media: authentic storytelling, emotional depth, and taste. She defines taste as “the ability to spot what’s working or what feels off – a critical gap that AI cannot yet bridge on its own”. And she is direct about the hiring implication: “For your startup to succeed, hire AI builders, who essentially become the managers of AI agents”.
Jaclyn Konzelmann of Google Labs makes the same point from a different angle: “Just because anything can be built, it doesn’t mean everything should be. Humans still have an amazing role to play in providing the taste, the vision, and the decision on what should be built”.
3) The Post-Keyboard World is Closer Than You Think
Grace Isford, Partner at Lux Capital, is watching the brain-computer interface space closely. Her view is that most people are still thinking about AI in terms of chat interfaces and image bots, and that this is a significant failure of imagination. “Very few understand how different our lives will be when generative AI is used in entirely new modalities”.
What she is pointing at is a world where you do not need to articulate what you want done, because the neural device will just know. That sounds far-fetched until you look at what is already in development. The compute implications alone are significant: new interfaces will require architectures built around smaller, more efficient models that can work in swarms, which she sees as the mechanism that will ultimately democratise access to AI.
4) Bi-Directional Media Will Replace the Feed
Darian Shirazi, Managing Partner at Gradient, makes a prediction that I think is underappreciated: “We’re entering an era of bi-directional media, one where we communicate with a model, and it consumes information from us and the world around us to provide very specific, personalised output”.
The implication is that the way we consume information is about to change completely. Not a feed that shows you what the algorithm thinks you want. A model that understands your world, your interests, and what is influencing you, and delivers content that is directly relevant to your life. He also argues that independent journalists will become the source of truth in this world, because legacy media brands will become less relevant as content is delivered through models rather than portals.
The report is free to download. If you are building, hiring, or investing in anything adjacent to generative media, it is worth an hour of your time.
Download the full report here
This piece first appeared in Talent Pools Nexus, our weekly read for leaders in research, insights and consulting. Subscribe on LinkedIn.
