Rideau-Rockcliffe Civic Notes
Index: 311 and artificial intelligence
A chronological list of posts on this blog whose main subject is Ottawa 311 or artificial intelligence, with article numbers, short summaries, and links.
Artificial intelligence
Unnumbered · 10 December 2023
Ottawa City Councillors need help
https://peterkarwacki.blogspot.com/2023/12/ottawa-city-zcouncillors-need-help.html
Councillors cannot absorb every brief, delegation, and meeting attached to a multi-billion-dollar city budget, and fatigue produces bad votes. The proposed remedy is an AI decision-support system loaded with source data plus explicit morals and ethics, used to augment rather than replace elected judgment. The historical model is Chile’s Project Cybersyn (1971–73), Stafford Beer’s operations-research network that let managers see economic data and send directives. The piece argues that no major city decision should proceed without an AI advisor’s recommendation. A version also ran on Unpublished as “It's time for AI in Ottawa City management.”
#114 · 17 February 2026
A musical interlude
https://peterkarwacki.blogspot.com/2026/02/a-musical-interlude.html
A short demonstration: lyrics by the author, music and vocals generated in Suno. The point is not the song. It is that generative tools already do work that used to require a studio, and that outside estimates put about half of entry-level white-collar jobs within reach of AI agents inside five years. Cultural production is treated as an early, visible example of a labour shift that will also hit administration, analysis, and municipal staff work.
#133 · 23 April 2026
Current Events forecast future trends
https://peterkarwacki.blogspot.com/2026/04/current-events-forecast-future-trends.html
Returns to municipal AI as an operating method, not a slogan. Road-safety and similar files are still decided on multi-year historical averages while current events move faster than the briefing cycle. Staff should be paired with better models so live data, not last year’s roll-up, informs the next intervention. Unused city data is the bottleneck: sensors and datasets already exist; the decision process does not use them in time.
#255 · 5 July 2026
The Yips in politics
https://peterkarwacki.blogspot.com/2026/07/the-yips-in-politics.html
Uses “the yips”—performance anxiety that wrecks a practiced skill—as a metaphor for council delay. AI is offered as the antidote because it does not freeze under political heat. Suggested uses include simulating thousands of municipal outcomes before a long debate, synthesizing briefing books, and allocating resources with an auditable rule instead of a closed-door bargain. Cybersyn appears again as proof that decision-support systems predate chatbots. The claim is not that software should vote. It is that fear and volume are currently doing the deciding.
Unnumbered · 20 July 2026
Robot Doggie PeeJayKay is going Canvasing
https://peterkarwacki.blogspot.com/2026/07/wake-up-robot-doggie-canvasing-is-just.html
A lighter AI piece. It distinguishes what language models can do—scripts, slogans, research—from what they cannot do: buy hardware, hold a bank account, or walk a street. It then compares 2026 quadruped options, including Sony Aibo, Unitree Go2, and cheaper wheeled companions, as a canvassing stunt. The serious thread is capability versus agency. AI is already useful for campaign content; physical presence still belongs to people and machines that someone has to purchase and operate.
#361 · 2 August 2026
AI Governance Issues
https://peterkarwacki.blogspot.com/2026/08/ai-goverance-issues.html
The main municipal-AI policy post. Ottawa is already using AI—staff prompts in the thousands, a résumé-screening tool, systems reading infrastructure footage—while the governance framework is still a promised 2026 document. The proposed standard is to publish an inventory of tools, disclose use by default, guarantee human appeal, test before deployment, and give the Auditor General an annual review. Adoption is ahead of oversight. A disclosure-and-audit rule has to exist before the tools outrun accountability.
#402 · 11 August 2026
Yes, Somebody is Looking
https://peterkarwacki.blogspot.com/2026/08/402-somebody-is-looking.html
A short traffic-and-audience note. Blog views are rising, and some “readers” may be AI bots rather than voters. Artificial intelligence is incidental here, but the post is one of the few that treats automated agents as part of the public information environment, not only as a city-hall tool.
#475 · 31 August 2026
Power to the people
https://peterkarwacki.blogspot.com/2026/08/power-to-people.html
AI appears as electricity demand, not software policy. Hydro Ottawa’s large-load queue includes data centres, some tied to generative AI, that would nearly double local draw if connected. The piece walks through unnamed projects, bill pressure, IESO unserved-energy forecasts, and the lack of a public day-ahead or ride-through standard. Related follow-up: #481 The Juice is worth the Squeeze.
311
#157 · 16 May 2026
PPP Pee Pee and Poo poo
https://peterkarwacki.blogspot.com/2026/05/ppp-pee-pee-and-poo-poo.html
A public-washroom and facilities post that treats 311 as the reporting channel that should make cleanliness and outages visible. Suggestions include QR codes in units that feed ratings and rapid-response tickets, plus integration with maps and 311 so complaints, usage, and surveys sit in one loop. 311 is the measurement system, not the headline topic.
#200 · 7 June 2026
FOIA to access Open Data?
https://peterkarwacki.blogspot.com/2026/06/foia-to-access-open-data.html
Residents should not need a freedom-of-information request to learn who owns a dataset or what the city already measures. The published scorecard includes 311 call-resolution times by category and by ward, so neighbourhoods can see whether service is even. 311 is framed as an equity and management metric: if it is not published in readable form, it is not being managed.
#244 · 30 June 2026
The Satirical Interview with an Ottawa Mayor
https://peterkarwacki.blogspot.com/2026/06/the-satirical-interview-with-ottawa.html
Satire on city statistics. One exchange says satisfaction surveys ignore people who stopped calling 311 years earlier. The 311 point is methodological: official service scores can be survivorship bias if the sample is whoever still bothers to phone.
#345 · 27 July 2026
Third Party news, not mine, not the city, not main stream
https://peterkarwacki.blogspot.com/2026/07/third-party-news-not-mine-not-city-not.html
The first clear public write-up of the 311 finding. A review of Ottawa’s open 311 dataset showed more than 123,000 service tickets bulk-closed in a six-day window, presented as a data-integrity problem rather than a flood or pothole story. The surrounding argument is that thinner local newsrooms make resident analysis of open data, MFIPPA files, and 311 extracts a substitute watchdog. Residents are told to pull the portal datasets and publish what they find.
#387 · 6 August 2026
There is a duty to consult but an imperative to lead
https://peterkarwacki.blogspot.com/2026/08/there-is-duty-to-consult.html
Contains the operational 311 proposal, often summarized as “Modify 311. Don’t build another app.” Instead of a new resident app, 311 itself should become GPS-aware: near an active capital job, the system should show project status and ask whether the new ticket belongs to that file. Linking at creation is offered as the fix for orphaned tickets and for the 123,000 bulk-closures. Engagement reform—agendas, disposition logs—sits in the same post. 311 is the enforcement layer for “the city already knew.”
#415 · August 2026
Enforce those bylaws!
https://peterkarwacki.blogspot.com/2026/08/enforce-those-bylaws.html
Smoking and vaping rules already exist; enforcement is mostly 311-driven rather than patrol-driven. The post asks for published 311 complaint volume and bylaw response times so residents can see whether complaints close or pile up. 311 is the evidence trail for a law that looks present on paper and absent in parks.
#416 · 13 August 2026
Whistle blowers — do we support them enough?
https://peterkarwacki.blogspot.com/2026/08/whistle-blowers-do-we-support-them.html
Uses the 311 bulk-closure, police-database auditing, and St-Laurent inspection findings as cases where someone inside likely knew first. Internal disclosure channels should be tracked the same way 311 tickets and Auditor General findings should be tracked: volume and outcome, not a black box. 311 is the analog for making a hidden process countable.
#427 · 16 August 2026
For the “How” — I’m your man
https://peterkarwacki.blogspot.com/2026/08/for-how-i-m-your-man.html
Applies the ADKAR change sequence to ward files. Under “Ability,” the concrete tool is a 311 rebuild: GPS project lookup and ticket-to-project linking at creation, using the system residents already call, so 123,000 orphaned tickets are not a recurring bulk-close event. 311 is treated as delivery infrastructure, not a call-centre slogan.
#443 · 21 August 2026
Lord, Give me a sign!
https://peterkarwacki.blogspot.com/2026/08/lord-give-me-sign.html
A method post. Pulling 311 data on West Rockcliffe flooding did not produce a simple flood narrative; it produced the six-day bulk-closure of about 123,000 tickets. The lesson stated is “evidence first”: start with 311 extracts, audited financials, and Auditor General reports, and accept the story the file actually contains.
Unnumbered · 7 September 2026
How to search for City Information
https://peterkarwacki.blogspot.com/2026/09/how-to-search-for-city-information.html
Not a 311 exposé. It places 311 and service requests inside Ottawa’s fractured digital map: live tools live on siloed domains while ottawa.ca search ranks old PDFs. Residents are told to bookmark ServiceOttawa, geoOttawa, DevApps, Recreation, Engage, and Open Ottawa, and to use site:ottawa.ca when they must search policy. 311 belongs in the “my services” cluster that the front door currently fails to find.
How the two threads connect
The 311 spine is: find the 123,000 bulk-closures in open data (#345); treat them as a data-integrity failure (#443); fix 311 in place with GPS project linking (#387, #427); publish resolution times by ward (#200).
The AI spine is: use AI as council decision support (2023, #255); use live data rather than stale averages (#133); put disclosure and audit rules on tools already in use (#361); watch the grid cost of AI compute (#475).
No comments:
Post a Comment