Despite what the companies said about every job changing, three jobs in a democracy will never change. We still need dutiful elected officials, dedicated civil servants, and informed voters.
On Monday, October 5, the City Council’s Committee of the Whole held a ten-hour hearing on the risks of artificial intelligence. The Speaker, Council Members, and the Public Advocate asked 218 questions, and 54 members of the public and expert witnesses testified, putting 161 concrete proposals on the record. If you want to re-run the counts, we made a transcript site and a short dataviz for you. (You could turn it into a project for Data Through Design 2027!)
As for what was said, the simplest summary is disappointment. Not in the Council, but in how the administration answered questions. The consumer and worker protection work sounds great, but it doesn’t sound like OTI has a technical team to back it up. With news that the City’s head of AI and machine learning is leaving, this was yet another Council hearing where OTI said it has yet to build out its algorithmic accountability team. And the statement that OTI has yet to see an opportunity for positive uses of AI in government overlooks strong examples from New Jersey, New York State, and around the world. (Read what the City said.)
I’m also proposing a public interest technology institute with CUNY, funded by taxing the AI companies.
For the allies in the room who have been fighting for accountability in automated decision-making systems in government (the term of art we used before AI), this was frustrating. If we can’t get concrete, values-based technology leadership inside this administration, there might as well not be a city office named “Technology” and “Innovation.” NYC needs an OTI that can deliver solutions for New Yorkers and upskill government staff. I really hope people inside the Mayor’s office know how precious this moment is. This is a unique moment to lead in public interest technology, and it will take more than a flashy media event on Coney Island.
Look, I know my lack of any college degrees keeps you from taking my résumé seriously, but call me. The gods know I’m cheaper than any accounting cartel, and I know how to commit, push, review PRs, and build a thoughtful AI harness, plus I have a solid understanding of policy.
As for the companies, they made it clear they will say what they need to say to look like respectful stewards of technology. Meta’s representative was one of the best-trained marketing representatives I’ve ever seen in a policy shop. They are also leaning hard into the “human in the loop” as a “meat proxy.” Nathalie Maréchal’s piece in Tech Policy Press is a real eye-opener.
Testimony worth reading
Here is written testimony from allies who have pushed for transparency in automated decision-making for years, and who have presented at Open Data Week and School of Data.
- Jordan Shapiro, BetaNYC associate board member and founder of JD Digitalization, who walked into the building with me. He asks the City to standardize on one contracted AI model with its own governance rules, have the Office of Algorithmic Accountability set and publish performance standards, and train agency coordinators to teach responsible AI use. – Read Jordan’s testimony
- Cynthia Conti-Cook, Director of Research and Policy at the Collaborative Research Center for Resilience. She traces a decade of warnings about automated decision systems, back to the City’s 2017 task force, and lays out how New Yorkers pay three times when harmful technology is procured. – Read Cynthia’s testimony (PDF)
- Julia Stoyanovich, Director of NYU Tandon’s Center for Responsible AI, in her personal capacity. “Disclose it. Prove it works. Let people contest it.” – Read Julia’s testimony (PDF)
- BetaNYC’s final written testimony, submitted October 8, with our bill-by-bill suggestions and appendix. – Read it online · PDF
Lastly, a huge thank you to the City Council staff and to our neighbors who testified. And thank you to Jiahao Chen for his service to the City. Jiahao, thank you for anchoring the City’s AI practices in a thoughtful knowledge of the technology. We look forward to seeing what you do next.
On Oct 23rd, we are organizing a BetaBuilders virtual event around civic AI tools for NYC’s democracy. We are looking for tools and skills that you can add to pipelines, workflows, and harnesses. If you know what I’m talking about, email me what you’ve got. noel+bb(at)beta.nyc. If you don’t, save the afternoon for a fun-filled demonstration of AI tooling.
The future is unwritten.
Noel Hidalgo
