← Team
Team · SVV

Barry Downes

Managing Partner

Founder turned investor with deep AI expertise

Barry Downes
Barry · Managing Partner

Barry founded Sure Valley Ventures in 2017 and leads it as Managing Partner. Before SVV, he started FeedHenry, the enterprise mobile platform Red Hat acquired in 2014, and was CEO of the Walton Institute (previously TSSG), an internationally recognised research and innovation institute. A computer scientist by background, he holds an MBA from UCD Smurfit School, completed the Venture Capital Executive Programme at UC Berkeley Haas, and is a Certified Investment Fund Director.

He writes AgenticInvestor Weekly on how AI agents are changing investing, and built the Agentic Investing Maturity Model that SVV runs on and publishes openly.

Writing

AgenticInvestor Weekly

Subscribe on LinkedIn →
AI is getting more capable. Markets are getting less forgiving.The Nasdaq fell 5.52% on Thursday. The S&P 500 lost 4.36%. It was the kind of day that makes even committed technology investors pause. Last week, I asked what the AI boom really costs. This week, markets became much less comfortable with the answer.Anthropic Opens the Books. What Does the AI Boom Really Cost?AI companies are selling more. Their models are getting better. So why are investors becoming harder to impress? This week offered a revealing answer. Anthropic opened its books. Google unveiled new technology and reported extraordinary cloud growth, then saw its shares fall sharply.The AI Apocalypse Is Cancelled* - Is the Black Box Finally Opening?What if AI’s black box is easier to open than we thought? That is the question I keep coming back to this week. We have spent years debating what increasingly capable models can do. A promising line of research is asking whether we can understand much more clearly how they do it.The AI boom meets the balance sheetYou can be right about AI and still make the wrong investment. This week brought that tension into focus. Concerns about financing the infrastructure build-out collided with another round of model launches. Meanwhile, developments in AI security raised questions about the cost of putting increasingly capable agents to work.When AI starts making discoveries, who captures the value?If AI can produce new ideas faster than experts can validate them, where does the investment opportunity move? That is the question connecting three developments in this week’s AgenticInvestor briefing: AI-generated mathematical research, Google’s enterprise agent launch and another reduction in the cost of intelligence.