How our polls work

Real participation. Visible limits. No invented results.

The question and answer options

If AI makes us more productive, who should benefit most?

1. Workers — Better pay or shorter working hours
2. Customers — Lower prices and better services
3. Businesses and investors — Profits and reinvestment
4. Society — Stronger public services and shared benefits
5. It depends — There is no single answer

Choose the group you think should benefit most. This is a question about preferences, not a test of factual knowledge. “It depends” is included because the answer may depend on context.

One question, two entry points

Drop Your Take and Rethinkist use the same question ID (ai-productivity-001) and shared Cloudflare D1 database. Results combine accepted votes from both websites. Answer order is fixed and may affect responses. There is no weighting, population adjustment or claimed margin of error.

Who takes part?

Visitors choose whether to participate. They are not randomly sampled. There is no demographic verification, identity check or evidence that they represent a country, workforce or the world. The unit counted is an accepted submission, not a verified unique person.

Integrity and abuse controls

Votes require a server-validated Turnstile check. Signed browser cookies, network-level limits, request limits and database uniqueness constraints reduce simple repeat submissions. Up to five accepted votes per network are permitted for this poll to accommodate shared connections. The cap can exclude legitimate participants and cannot stop every attempt to vote repeatedly. Browser identifiers are separate on the two domains. Neither mechanism proves “one person, one vote”.

Reading the results

Counts start at zero. There are no example votes, seeded popularity figures or simulated participants. Each answer is shown with its count and its share of all accepted votes. Percentages are rounded and may not total exactly 100%. The result display includes the latest retrieval time. A database error is shown as unavailable, never as zero participation.

Dates and changes

The question is prepared for launch on 8 November 2026. Collection begins only when the public voting service opens. No closing date has been set. If the wording or answer choices materially change, we will use a new question ID instead of mixing different questions. Test submissions belong in a separate test database, never in published totals.

Sharing and measurement

Sharing is optional. With analytics consent, a native share action or copied link is counted separately from a visit through a share link. These are aggregate actions, not verified deliveries, conversations, unique referrals or additional votes. No reward or payment is offered for recruiting others.

Corrections and data use

Contact talk@dropyourtake.com about errors or suspected abuse. Any later removal of invalid votes or substantive methodological change will be disclosed. Reuse must state the question, participation count, retrieval date and self-selection limitation.