The Investor's Edge Lives in Their Head — Anekha Sokhal, Moshi

Highlights

• The world generated approximately 175 zettabytes of data in 2025.

• The human brain operates on about 20 watts of power.

• Financial edge consists of seeing information first, understanding blind spots, and having conviction to size trades.

• Moshi processes 2.4 billion data points and learns 300 new things every four seconds.

Timeline

1. 00:00 - 03:35

   Data Overload and the Financial Paradox

   Modern financial professionals are overwhelmed by information streams spanning newspapers, PDFs, podcasts, and social media platforms like X and LinkedIn. More data does not translate to better information, which widens the gap between available data and human attention spans. Despite this disparity in data consumption, the human brain remains vastly more energy-efficient than high-powered computing infrastructure.

   • Financial investors face a massive paradox where data volume increases exponentially while human attention remains flat.
   • The world produced 175 zettabytes of data in 2025.
   • Human brains operate efficiently on about 20 watts of power, whereas imitation machine systems require entire power plants.
2. 03:35 - 06:16

   Defining Edge and the Limits of Replacement AI

   Any investor seeks market edge through unique insights, risk awareness, and position sizing. Because competitors use identical models and prompts, technical tools alone fail to create an edge. True expertise accumulates over decades of mentorship and industry experience, meaning that intellectual property leaves the firm when an investor departs.

   • Market edge relies on knowing developments early, recognizing second and third-order derivative blind spots, and holding trade conviction.
   • True institutional edge cannot be downloaded via generic models or prompts because it resides within experienced investor intuition.
   • AI tools often attempt to replace human judgment instead of supporting the core reasoning engine.
3. 06:16 - 10:44

   Infrastructure Reality Versus Weekend Dashboards

   While AI tools encourage single-person pods to replace entire quant stacks, scaling those tools introduces major hurdles regarding API credits, data storage, and infrastructure security. Financial trading involves physical cargo and real corporate clients, meaning autonomous errors cannot be resolved with simple apologies. Separating the computational grind from human judgment optimizes workflow efficiency.

   • Building a weekend dashboard differs significantly from deploying production-grade systems that require security, storage, and reliability.
   • Financial firms cannot hand decision loops over to autonomous agents because accountability requires a human to justify actions.
   • Machines should handle data collection and cleaning while humans retain responsibility for final judgments.
4. 10:44 - 16:20

   Scaling Institutional Analysis with Moshi

   Moshi bridges the gap by ingesting massive market data feeds alongside private subscriptions and emails, pushing relevant context directly to traders. By handling data gathering and identifying blind spots, the platform frees traders from administrative distractions and conversation loops with static language models. This infrastructure delivers institutional-scale data science capabilities to individual desks without requiring dedicated internal engineering teams.

   • Moshi functions as a 24/7 quantitative analyst powered by six deterministic agents tracking market data and proprietary investor documents.
   • The system processes over 2.4 billion data points while continuously measuring hallucinations and verifying source claims.
   • Institutional-grade analysis increases operational efficiency by up to 35 percent and provides deep contextual awareness.