What’s the context? Sivesh Sukumar, a Principal on Balderton Capital’s investment team, returned to X in August 2026 with a short reflection on the trade-off between safety and utility in artificial intelligence. He described AI as “probabilistic software”, arguing that unknown edge cases will always exist and can be particularly consequential for incumbent organisations with substantial existing exposure.
Why it matters: Sukumar’s post frames uncertainty not only as a technical and governance challenge, but also as a potential opening for startups. His observation is that incumbents with more to lose may be slower or more cautious in deploying systems whose behaviour cannot be fully predicted in advance, creating room for new companies to build products, controls and workflows around emerging AI capabilities. The post does not announce a transaction or identify a specific startup; it is a concise investment-oriented observation about where risk, adoption and entrepreneurial opportunity may intersect.
Profile context: Sukumar joined Balderton’s investment team in 2022 after working in the fintech team at JP Morgan, where he looked at strategic investments, acquisitions and partnerships across the investment bank, and spent time in securitised products on structured-finance opportunities in Europe. Earlier internships included trading at Morgan Stanley, business development at Finimize, and software engineering at Vivacity Labs. His background includes software engineering and machine learning.
Read more: Sivesh Sukumar on X · Balderton Capital
Leave a Reply