In an era defined by exponential technological leaps, a quiet but fundamental shift is brewing inside venture capital. The conventional dynamics of investment, such as personal networks, slow diligence cycles, and pattern recognition rooted in past successes, are beginning to fray under the weight of a new force: artificial intelligence.

Few are voicing this tension as directly as Eldad Tamir, CEO of FINQ, an AI-powered asset management firm. In a recent viral LinkedIn post that garnered considerable traction among technologists and investors alike, Tamir laid bare what he describes as the “very complicated relationship” between venture capital and artificial intelligence.

“VCs excel at identifying new tech that fits into existing markets,” he wrote. “But AI doesn’t fit that model… It’s not iteration—it’s reinvention.”

Tamir’s critique isn’t abstract. He argues that many investors are comfortable funding AI when it presents itself as a “tool”, like a plugin for productivity or a co-pilot for coders. But AI as a foundation, one that replaces human fund managers or reimagines entire industries, is often too radical for traditional frameworks.

That conservatism may already be costing the venture capital world its edge.

The QuantumLight Experiment

One of the boldest counterpoints to this inertia is unfolding in London, where Nik Storonsky, founder of the $45 billion fintech giant Revolut, is applying an engineer’s mindset to the venture capital (VC) industry itself.

Storonsky, alongside Ilya Kondrashov, launched QuantumLight in 2022 with a provocative thesis: eliminate human subjectivity from venture investing entirely. The platform scrapes more than 10 billion data points and tracks over 700,000 venture-backed companies globally, identifying high-potential startups using algorithmic pattern detection.

Their results are difficult to ignore. All 17 of QuantumLight’s current portfolio companies were sourced entirely by AI with no pitch meetings, no introductions, no gut instinct. According to early reports, these companies are outperforming top-tier VC portfolios by 2x. Storonsky’s conviction runs deep: he personally contributed 25% of QuantumLight’s recent $250 million fundraise, joined by billionaire founders and institutional capital.

“If we removed humans from flight navigation, factory floors, and cancer diagnostics, then why not in investment decisions?” Kondrashov recently asked in an interview.

It’s a compelling question, especially when paired with new data from PitchBook indicating that 30% of VCs now use AI to source at least half of their deals, up from just 7% in 2019. Still, only a fraction of firms fully trust AI as the decision-maker.

Reimagining the Model

Tamir and Storonsky approach the AI revolution from different angles: one as a capital allocator, the other as a founder-turned-VC disruptor. However, both agree: traditional venture models are increasingly out of sync with the pace and nature of innovation.

This disconnect has real implications, especially as entire categories of startups emerge that operate outside human legibility. AI-first biotech platforms, autonomous financial engines, or decentralized decision-making protocols are harder to evaluate through the lens of “founder charisma” or “network adjacency.”

FINQ, Tamir’s firm, is leaning into this paradigm shift with its upcoming line of AI-powered ETFs, built not on sector nostalgia or index mimicry, but on machine-led rebalancing and dynamic asset mapping. Though details are under wraps, it represents the same philosophy at play in QuantumLight: trust the math, not the mythology.

As venture returns compress and competition intensifies, some believe that embracing algorithmic logic may be the only way forward. According to data from Cambridge Associates, the top quartile VC funds from 2010–2020 have seen IRRs decline by nearly 40%, suggesting that traditional alpha is harder to come by in today’s crowded landscape.

The Meeting of Minds?

What’s perhaps most fascinating is how Tamir and Storonsky are articulating a shared vision without coordination: one in which capital no longer moves at the speed of trust, but at the speed of data.

Both believe that the next generation of fund performance will come not from who you know, but from what your systems know. And in a world increasingly dominated by models that learn and adapt faster than humans can convene a Monday partner meeting, the argument for retooling venture capital becomes hard to ignore.

One wonders what would happen if these two pioneers sat down to compare notes. It might just spark the blueprint for venture’s algorithmic future.