Aiden Livingston is the founder of Casting.AI and the author of the upcoming book, “Build a Bot Workshop — A Guide For Creating The Next Generation of Ai Startups”. For a counterpoint, see my more-skeptical note at the end.
Ever wonder why we vote for President on the first Tuesday in November? It seems to defy any conventional wisdom on how to get maximum turnout. There’s a reason for that, it was designed to accommodate farmers’ needs over a hundred years ago. The reasoning was Sunday was the Sabbath, so that was a no-go. Most people lived on farms, many of them remote, so we gave them the whole day of Monday to allow for painfully slow animal-propelled travel, so they could then cast their vote on Tuesday.
Now in 2016, less than 1% of the population are farmers, cars have existed for more than a century, and when most young people hear the word “Sabbath” they are more likely to think of Ozzy Osbourne than church on Sunday. Yet we still vote on Tuesday. why?
I have been designing digital products for over 5 years, including my latest project to create the world’s first AI-based talent agent called Casting.AI (watch your toes, I am dropping shameless plugs). The main lesson I have learned is the user is never wrong, your system is wrong.
It is flabbergasting to watch how far pundits will go to blame the voters for the problems with the voting system. Fully 28% of voters don’t vote because they report being too busy. Which seems a pretty reasonable excuse for not being able to wait in line for three hours in the middle of a workday.
The idea becomes all the more inane when you consider what they are lining up for, to use a touch screen to register their vote. Most people have upwards of 5 touch screens laying around their house, but we all lined up to use this one “special” screen, I suppose because we are hoping to give contact-driven flu infections ahead start on the season?
The justification may be, we need the staff to verify your identity with your ID. After all who could possibly crack the infallible system of photo identification? Aside from every person aged 16-20 in America — incidentally, my fake ID was flawless.
Somehow we all understand and appreciate the complexity and security of online verification processes when it comes to your life savings, but feel the same multi-billion security system pales in comparison to an elderly volunteer ability to look at your driver’s license.
It is beyond the extent of this short rant, but the system developed by virtual currencies like Bitcoin would make voting fraud all but impossible. If you are interested in learning more, or in need of a powerful sleep aid, look up the inner workings of the blockchain. We could anonymize parts of the voter’s ID so that each user could see their own vote correctly logged in the ledger. Ending fears of voter fraud is literally just that easy.
That is the main point, the problems that people have with voting are easily solvable. In fact, almost all of them could be solved by implementing existing solutions that have already been perfected by countless other online properties.
Letting people vote from their phones would increase security and eliminate the potential for voter fraud. It would also allow people to take more time when voting, rather than just voting for president then picking the down ballot candidates by whose last name they like the most. We could add a “more info” button for each candidate where they could in their own words explain their key positions. Why shouldn’t voting be a leisurely process of considering each down-ballot vote carefully in the comfort of your own home, not something rushed-through in what looks like a changing room in a school cafeteria?
One of Google’s top innovators, Jake Knapp, recently wrote a book called “Sprint,” which laid out a method many startups use to create the technology we love and use every day. The idea is simpl: identify possible improvements to your system and test solutions to make your product a little bit better each week.
The system hinges upon user feedback, from which you form new ideas based on user feedback and evaluate the effectiveness of your change based on more user feedback.
So today as we watch the spectacle of squeezing a 21st century electorate through a 19th-Century voting system, I hope that we can collectively take a moment to consider how can we start to implement all this new technology. Our goal should be to make the voting system better and not how can we better incentivize voters to wait in a three-hour voting line in 2020.
Ed. note: I’m less sanguine than Aiden about the prospect of voting-by-phone in the near term (though I’ll note a Larry Niven story from the late ’60s in which he describes half of a future Earth population “bothering” to phone in their votes). Voting doesn’t happen often, so we don’t have as many opportunities for that feedback loop he talks about. And without a paper trail, if something DOES go wrong (for instance, if the system gets hacked), we have nothing to fall back on to determine who actually won — and we’ve seen what happens when people start claiming the system is “rigged”. Someday yes, but without infallible voter identification and some way of ensuring security at the network level, I can’t see this happening yet. – cpd
Hey Colin, always a pleasure my friend! Your concerns about hacking and needing a paper trail are certainly shared concerns about going digital with voting. Perhaps in another piece, it would be interesting to explore just how the inner workings of blockchain technology can solve this. Because truly nothing is more tempting to hack than money, so when designing the technology they created a system that is impossible to hack, where the ledger is essentially a long record that is routinely time-stamped at exact intervals. Such that if you wanted to change even one vote you would need to change the entire sequence dating back to the beginning of the blockchain. Would be great if someone who understands blockchain technology and can explain it better than me could chime in (cough cough hint hint tech readers).