This Billionaire’s AI Was Supposed To Speed Up Policing. It’s Not Going Well.

Key Concepts:

• AI in policing
• Surveillance technology
• Data unification
• Project Sherlock
• Usability issues
• Implementation delays
• Bureaucratic obstacles

Project Sherlock: A Case Study in AI Policing Challenges

1. Project Overview and Goals:

• Project Sherlock: An AI-powered surveillance system designed to unify police data from 15 agencies in San Mateo County, California.
• Objective: To accelerate police investigations and improve situational awareness by analyzing data from license plate readers, surveillance cameras, arrest records, 911 calls, and historical police databases.
• Technology: Developed by C3AI in partnership with Amazon Web Services (AWS).
• Funding: Supported by a $4.5 million donation from Tom Siebel's foundation for range refurbishment and $12 million invested in the project.

2. Key Players and Involvement:

• C3AI: A $2.44 billion market cap company founded by billionaire Tom Siebel.
• Tom Siebel: Deeply involved in the project, including writing contract sections and hosting San Mateo cops at his offices.
• San Mateo County Sheriff's Office: Organized a "day of fun" at the Coyote Point Shooting Range to celebrate the project's progress.
• Amazon Web Services (AWS): Partnered with C3AI in the development of Project Sherlock.

3. Implementation Issues and Delays:

• Timeline: The planned rollout aimed to have 14 agencies live with Sherlock by January of the current year.
• Usability Issues: The system has faced challenges related to usability and feature oversights.
• Delays: Significant delays in implementation, in some cases exceeding a year.
• Agency Access: Seven agencies (Atherton, Belmont, East Palo Alto, Foster City, Menlo Park, San Bruno, and South San Francisco) reported not having access to a fully functional product and therefore hadn't seen any investigational benefit.
• Conflicting Reports: The San Mateo County Sheriff's Office claimed that Foster City and San Bruno police departments did have access along with Daily City, Redwood City, and San Mateo.
• Limited Beta Version: San Bruno's former interim chief, Susan Manheimer, stated the department was using a limited beta version of Sherlock.
• Developmental Stages: Foster City's Chief Cory Call indicated they were "not seeing any benefits yet as this project is still in the developmental stages."

4. Perspectives on the Delays:

• Menlo Park Police Department Chief David Norris: Stated that all 14 agencies were "behind the implementation schedule in the timeline" and that "While the technology itself is onboarded, the practical implementation has been delayed."
• Belmont Lieutenant Clyde Hussie: Confirmed the delays and stated, "I wish I had a good answer for you." He later added that some select staff were recently given access.

5. Key Arguments and Evidence:

• Argument: Despite significant investment and technological backing, Project Sherlock has struggled to deliver on its promises.
• Evidence: Delays in implementation, usability issues, and the fact that many agencies have not seen any investigational benefit.

6. Notable Quotes:

• Chief Cory Call (Foster City): "Not seeing any benefits yet as this project is still in the developmental stages."
• Chief David Norris (Menlo Park): "While the technology itself is onboarded, the practical implementation has been delayed."
• Lieutenant Clyde Hussie (Belmont): "I wish I had a good answer for you."

7. Synthesis/Conclusion:

Project Sherlock exemplifies the challenges of implementing AI in policing, even with substantial resources and technological expertise. The project's struggles highlight the bureaucratic obstacles and performance issues that can hinder the deployment of advanced surveillance technologies. Despite the initial promise of faster investigations and improved situational awareness, the full realization of these benefits in San Mateo County remains elusive.