Keynote Speech
summary
Claims that machine-learning models will be trained on public blockchains plus exchange logs to turn transaction graphs into identifiable lives, using existing open and closed LLMs and diffusion tooling. Evidence: data already exists on on‑chain + exchange off‑ramps and large banks; timeline given is tests within ~3 years and routine use in 5–10 years. What gets built is automated anomaly detectors that spit alerts to compliance or intelligence desks, and the result is rapid erosion of transactional privacy.
AI-written summary · about
chapters (10)
- 0:00The State of the Industry Update
- 3:56Concentration of Resources and Intermediation
- 7:10Systemization of Technologies
- 10:21Building a structure around them to prevent monopoly formation and tax avoidance
- 13:45Metadata as Activity Records
- 16:57The Potential Risks of Transaction Monitoring
- 20:10The vulnerability of enormous data sets
- 23:12The Changing Dynamics of Economy and Politics
- 26:25The Need for Cooperation and Working Together
- 30:07Hyperbitcoinization
related talks
46mKeynote: How Can We Trust Machine Learning?
50mKeynote Panel: AI in Finance
21mTowards Building a Responsible Data Economy
26mWhen Crypto Becomes Intelligent
17mKeynote: Breaking Boundaries - Blockchain as a Catalyst for Social, Environmental and Economic Scale
20mBlockchain, AI and Social Payments
06mData Privacy Concerns and Solutions When using LLMs in Production
44mAdvancing Cybersecurity with AI and Machine Learning: Opportunities and Challenges