Conference Presentation, Fireside Chat, Interview
How Data Became the Most Indispensable Resource for AI | Or Lenchner, Bright Data | RAISE 2026
- Anticipates data demand from the AI industry will grow faster and stronger than previously modeled, necessitating two full stack re-architectures and a shift toward visual data formats beyond text.
- Predicts robotics will become the dominant industry trend in 2026, driven by immediate needs for compute and data before public prototype releases, with a surge in visual data collection expected over the last 12 months to support industrial and humanoid operations.
- Foresees industrial robotics applications advancing faster than non-industrial ones due to well-defined use cases and a lack of public trust barriers, while non-industrial humanoids will initially perform simple tasks like laundry or dishes before general household deployment.
- Notes that early adopters, specifically in China, may see humanoids in homes next year, with complex tasks and full "physical AI helper" capabilities expected as failure tolerance in the real world drops.
- Plans to generate new ad hoc visual data sources via existing human platforms to address niche needs unmet by web data, reflecting a shift from egocentric views to human-to-human and robot-to-robot interaction scenarios.
- Expects traditional tech firms to pivot to AI for full-context answers, increasing requirements for talent, compute, and data, while Chinese companies focus on building bare metal data infrastructure and Western firms outsource non-core data sourcing to providers.
- Reports a previous annual run rate of $300 million and a surpassed mid-year target of $400 million, with plans to maintain focus on building the data stack without pursuing a short-term IPO.
- Cites a successful legal conflict with Bright Data as critical for the existence of current web data, AI model training capabilities, and the viability of web-searching chatbots.
- Maintains that AI cannot revert to a pre-AI state and that underlying technology will persist despite potential stock market corrections from extreme valuations.
- Warns that high startup valuations may limit employee upside, suggesting industry builders will likely continue working and wait approximately two years for the next market cycle.