Angel Investors Shouldn't Pay for Dreams

A Viral Post Sparks an Old Debate

A now-deleted post on Xiaohongshu — China's equivalent of Instagram — recently ignited a fierce debate in the venture capital world: 'Angel investors in China all come from hell.' The anonymous author argued that startups completing angel rounds almost always rely on personal connections rather than merit, and that VCs who dare to invest at the earliest stages barely exist.

The post was inflammatory and oversimplified. But it struck a nerve — particularly as ChinaVenture (投中网), one of the country's leading investment media outlets, was surveying industry sentiment ahead of its 20th Annual China Investment Conference.

Why This Matters Beyond China

The frustration is not uniquely Chinese. Across Silicon Valley, London, and Berlin, angel investors face the same existential question: why should I pay for someone else's dream?

In 2024 and 2025, early-stage AI funding has reached unprecedented levels globally. According to PitchBook, global angel and seed-stage deals in AI totaled over $12 billion in 2024 alone. Yet the failure rate for angel-backed startups remains stubbornly high — roughly 90% never return capital to investors.

The tension is particularly acute in AI, where:

• Compute costs can burn through a seed round in months
• Large incumbents like OpenAI, Google, and Meta can replicate features overnight
• Talent acquisition at the earliest stages requires premium salaries
• Regulatory uncertainty in the US and EU adds risk layers that didn't exist 3 years ago
• Revenue timelines for foundation model startups stretch far beyond traditional SaaS benchmarks

The 'Background' Problem Is Universal

The deleted post's core claim — that angel rounds go to founders with the right 'background' — resonates globally. Research from Harvard Business School has consistently shown that warm introductions account for the vast majority of funded deals in the US market.

In AI specifically, the pattern is even more pronounced. Founders from Google DeepMind, Meta FAIR, or OpenAI routinely raise $5-10 million seed rounds on reputation alone. Meanwhile, technical founders without elite pedigrees struggle to get meetings, regardless of their product's quality.

This creates what some call the 'credentialism trap' — investors use background as a proxy for de-risking, which reinforces existing power structures rather than discovering genuine innovation.

The Counter-Argument: Risk Has a Price

Experienced angel investors push back on the 'dream funding' critique. Jason Calacanis, one of Silicon Valley's most prominent angels and an early Uber backer, has long argued that angel investing is fundamentally about pattern recognition, not charity.

'You are not paying for a dream,' the logic goes. 'You are pricing risk based on available signals.' Those signals inevitably include a founder's track record, network, and institutional backing.