本案强调人工智能在早期投资中的比例性作用。InReach Ventures 自 2015 年起以机器学习取代传统拓展模式，其累计约 €65 million 投资中约 €50 million（约 77%）由 AI 主导，其第二支基金的项目获取比例为 100%。典型 VC 模式需依赖高频城市移动与偶然接触，而该公司通过自动化数据搜集与分析降低此类随机性，并完全取消分析师与助理岗位。其逻辑为 20 项投资中只需 1 项规模化，即可产生显著 carry，但回报周期可延长至 10 年以上。创始人预期未来单人基金规模可扩至 €1 billion，仍以 AI 为核心驱动力。
争议集中于人才入口被削弱与机器判断的不确定性。行业原有的职业梯队因自动化而被移除，导致无运营经验者失去进入路径。Entrepreneurs First 推出两年制项目以填补初级投资与创业培训空缺，强调判断力与品味无法完全替代。实践中，AI 仍出现识别错误，包括对 Bonanzinga 口音的误转写。他在创立 InReach 的第一年集中清洗数据，并持续监督 AI 决策，将错误建议标记为无价值。
尽管如此，执行结构高度清晰：AI 提供项目来源，人类负责最终筛选与关系建立。随机性被压缩，但完全依赖仍需谨慎。AI 可在规模化搜寻中提供显著效率，却无法承担长期投资关系管理的核心人类职能。InReach 的模型验证了一个趋势线：在资金部署比例与流程自动化程度迅速攀升的背景下，人类在价值链中的位置正向更上游的判断环节收缩。
The case highlights the proportional role of AI in early-stage investing. Since 2015, InReach Ventures has replaced traditional sourcing with machine learning, with about €50 million out of €65 million (≈77%) driven by AI, and 100% of its second fund’s deals sourced by artificial agents. Conventional VC work depends on high-frequency travel and chance encounters, while InReach reduces such randomness by automating data collection and analysis and eliminating analyst and associate roles. Its model assumes that one out of every twenty investments must scale to generate carry, with return cycles extending beyond a decade. Bonanzinga projects that future solo investors could manage €1 billion funds supported primarily by AI.
Criticism centers on removed entry paths and uncertain machine judgment. Automation eliminates the traditional talent pipeline, excluding those without operating backgrounds. Entrepreneurs First launched a two-year fellowship to fill junior-level training gaps, stressing that judgment and taste remain non-automatable. In practice, AI still commits errors, including mis-transcribing Bonanzinga’s accent. He spent his first year building clean datasets and continues to supervise AI decisions, marking incorrect suggestions as uninteresting.
The operational structure is unambiguous: AI sources deals, humans perform final filtering and relationship building. Randomness is compressed, but dependence on automation remains partial. AI achieves scale in sourcing but cannot replace long-horizon relationship work. InReach demonstrates a trendline: as capital deployment and automation ratios rise, human contribution shifts upstream into higher-order evaluative roles.