The headlines of AI sell ads, but do they impact your business?
Looking for different takes on AI, that matter?
The headlines of AI sell ads, but do they impact your business?
The data is eye‑catching. OpenAI now represents 45% of Microsoft’s $625 billion commercial backlog.
The hook is just as dramatic: while 74% of firms are increasing AI budgets, Gartner predicts 40% of agentic AI projects will be canceled by 2027 due to unclear value.
The take: PwC suggests 2026 is the year agents move from demos to industrial‑strength, a sharp contrast to the 95% AI failure rate cited in the MIT study discussed in our first episode.
Do these takes impact your business?
Do they shape your AI strategy, or expose where AI may not deliver value?
Why We Launched Three Takes on AI
Campbell Robertson, Michael Muhlfelder, and I, Brian Silverman, kept coming back to the same question:
Are the headlines, hype, and AI news of the day actually helpful for IT and business leaders trying to create value, and avoid waste?
That question led to the launch of Three Takes on AI, a podcast built on a simple belief: one take is not enough. Perspective, experience, and choice are essential in navigating the AI landscape.
· Campbell brings the lens of trust, governance, and responsible adoption.
· Michael focuses relentlessly on ROI and the return on hard‑earned capital.
· I aim to balance technology possibility with real business strategy, AI for outcomes, not hype.
Our podcasts are avilable on YouTube, with additional background and reference links available on our website. For those on the move, we’re also available on Spotify.
We launched in November 2025, releasing new episodes every other Tuesday.
Podcast Summaries: (Link to podcast below)
Shadow AI: From Risk to ROI
While headlines focused on the MIT study’s grim ROI statistics, our discussion surfaced a critical counterpoint: Shadow AI is often where early value is already being realized.
Insight: We emphasized that Shadow AI is not just a compliance issue, it is a signal of unmet demand. Organizations that study why employees adopt AI tools informally gain a roadmap for sanctioned, higher‑value AI deployment.
The AI “Bubble”: Still Growing or Ready to Burst?
We explored investor anxiety and market froth around AI, a conversation that feels even more relevant today.
Insight: Beyond capital markets, we discussed the organizational bubble: leaders funding AI initiatives without clear ownership, success metrics, or accountability, creating internal versions of the same risk investors fear.
Getting ROI from AI, or Pouring Money into the Hype Machine
This episode zeroed in on Michael’s core question: Where is the money?
Insight: We differentiated cost avoidance, productivity lift, and revenue creation, noting that many AI programs fail because they mix these outcomes together instead of choosing one primary value thesis.
Three Takes: The 2025 Year‑End Wrap‑Up
We closed the year reflecting on what energized us about AI and what concerned us.
Insight: A recurring theme was that organizational readiness, not model capability—was the dominant constraint in 2025. Talent, process change, and leadership literacy mattered more than access to cutting‑edge models.
Our Take: Agentic AI at a Crossroads — Why 2026 Matters
Campbell framed 2026 as a make‑or‑break year for agentic AI, a view we all discussed.
Insight: We highlighted that the risk isn’t agents failing technically, but failing organizationally, when autonomy outpaces governance, change management, and human trust.
Bringing AI Out of the Shadow
In our most recent episode, we revisited Shadow AI from a more mature perspective.
Insight: We discussed practical pathways to move from unmanaged experimentation to intentional enablement, including open communications, policy guardrails, approved tooling, and executive sponsorship, without killing the innovation that made Shadow AI valuable in the first place.
Looking Ahead
I’m proud of how these conversations have evolved, from three distinct viewpoints into a shared, experience‑driven dialogue shaped by our experience across past technology waves: client‑server, internet, mobile, and cloud.
Our next podcast (February 10, 2026) will focus on Human vs. AI intelligence:
· What truly differentiates human intelligence from AI?
· Why does that distinction matter for organizations?
· How should leaders plan for humans and autonomous agents working side by side?
Sources & Background
· Microsoft’s $250 Billion Problem Has a Name: OpenAI
· Agentic AI’s Great Divide: Hype Meets Harsh Organizational Realities
· 2026 AI Business Predictions
· Campbell Robertson – LinkedIn