Recently, Councillor Lisa Robinson called me out on the publication of my piece, “PROJECTION VOTES (2026).”
The Councillor’s criticism of the fairness, implications and justification of the piece may be correct.
I published the piece as a guesstimated campaign snapshot based on AI research. However, Councillor Robinson’s criticisms have made me reconsider the merits or appropriateness of the piece.
Her questioning of this publication may be debatable but if there is any risk of tainting the campaign or swaying voters’ views based on hearsay or erroneous information, then the piece should be withdrawn. Though voters should not be swayed by ‘hearsay,’ such a piece could have unintended negative implications. It should definitely be withdrawn for its errors.
The philosophical question
In light of the possibility of misleading or misdirecting the electorate, there is a philosophical question to consider, the pro’s and con’s of publishing such a ‘guesstimate’ piece.
PRO’s
Transparency: I labelled it a “guesstimate,” not fact and explained that this work was AI based (two AI models published misleading information regarding the position of the Cornell result in the mayoralty race.)
Consistency: My projections covered all the races, not just cherry-picking any particular one or one specific constituency. So there is universality and equitable fairness in the piece.
Engagement: Data visuals like this are the kind of content readers actually share and read as it is significantly different than the typical text-only local coverage.
CON’s
Conflict of interest optics: publishing “projections” during an active campaign. Even with disclaimers, it can look like I’m using my own outlet to shape perception of who’s viable and who isn’t, an intent I never had. The possibility of such influencing of voters needs to be considered and a review of contining the publication of the piece undertaken.
Accuracy risk: The Cornell anomaly proves the model can be badly wrong. If a projection turns out embarrassingly off, it undermines credibility right when it matters most. Using only two AI models may have been too narrow of research tools. I need to broaden my research.
Invites scrutiny: Other candidates or voters could frame this as campaign material dressed up as journalism, opening me to complaints about fairness or platform misuse rather than a simple examination of campaigning numbers.
Bottom Line
The merits and the downsides of running this piece need to be seriously reviewed and a decision about whether to continue or terminate it needs to be made. Satisfying a personal itch of curiosity does not justify publishing a possibly misleading and dubious piece.
