CB Insights AI 100 and Sustainable Innovation

CB Insights just unveiled their 2025 AI 100, full of exciting names. With the landscape shifting quickly, how do we separate sustainable innovation from hype?

The rapid evolution is clear – AI agents, security, observability, and governance are taking center stage in 2025. This dynamism means looking beyond today’s hot categories is crucial. For me, two areas stand out as critical for sustainable advantage:

(1) Vertical AI’s Deep Advantage - Embedding deep domain expertise to navigate industry nuances, regulations, and build refined, sector-specific datasets – the kind of quality data crucial for reliable AI – offers a powerful defense against generalist models. This leads to clearer ROI and potentially stronger moats, and the funding momentum here is telling.

(2) Evolving Data Strategies - AI runs on data. While foundational infra like vector DBs remain key, the fluctuating focus on Synthetic Data (Aaru ‘25 vs. Hazy/Mostly AI ‘23) raises questions. With real-world data facing limits (quality, privacy, access), is the strategic importance of synthetic generation being fully appreciated to fuel future model advancements, especially as more advanced reasoning models might be needed to generate truly useful synthetic data? Perhaps 2026 is its year.

What’s your perspective? Will vertical or horizontal solutions dominate? Will synthetic data adoption take off in the coming year?

Originally posted on LinkedIn. Join the discussion and share your thoughts there.