Forecasting Tomorrow’s Technology Today

As it’s been said, prediction is very difficult, especially about the future

3 min read

Forecasting Tomorrow’s Technology Today
Photo: Volker Möhrke/Corbis

Early in 2014, IEEE Spectrum teamed up with SciCast, the “Bayesian combinatorial prediction market” group based at George Mason University, in Fairfax, Va. And when our January Top Tech 2015 issue hit the Web, IEEE Spectrum added something new to a few of its articles: the opportunity for readers to participate in IEEE Spectrum SciCast forecasting and match wits with experts by making their own predictions about the future of technology.

SciCast founders Robin Hanson, Kathryn Laskey, and Charles Twardy built the system to allow large numbers of forecasters (some 10,000 have signed on so far) to collectively prognosticate on technological progress. Initial support for SciCast came from the U.S. Intelligence Research Projects Activity.

The publication of our annual prediction issue seemed an ideal time to introduce Spectrum readers to SciCast. So, for example: Rachel Courtland’s piece, “The XPrize’s Lunar Deadline Drifts,” reported that the Google Lunar XPrize team had extended the prize deadline to 31 December 2016. On SciCast, we ask two questions: “When will a spacecraft land on the moon and fulfill the requirements of the US $20 million Google Lunar XPrize Grand Prize?” and “Which team competing as of December 2014 will win the Google Lunar XPrize before December 31st, 2016?

Considering Erico Guizzo’s “A Robot in the Family,” we ask, “When will an in-home robot—costing less than $100,000 and capable of performing at least three household chores—be available to the general public?” At press time, SciCast predictors say there’s a 61 percent probability that automated maids will arrive before 2022 does.

The term “prediction market” sounds exotic; the thing itself isn’t. To see one at work, look no further than the pari-mutuel tote board at almost any horse track. In pari-mutuel betting, of course, the odds offered are basically the ratio of the total amount bet on all horses to the amount wagered on one particular nag. The process is dynamic, as bettor-forecasters factor group consensus into their own calculations.

Prediction markets can seem spectacularly prescient, as the Dublin-based Web trading exchange Intrade showed with its spot-on forecasting of Barack Obama’s U.S. presidential victory in 2012. At Intrade, however, members put up actual cash to buy “contracts” on specific event outcomes, which caused big problems with the U.S. Commodity Futures Trading Commission [PDF], leading to the company’s demise.

SciCast forecasters invest points, not cash. And if the augurs augur well, they win more points, bragging rights, and sometimes more tangible prizes offered by SciCast and its sponsors. And, in a change from most prediction markets, SciCast allows direct bets on a variety of different outcomes and permits bettors to invest points to assign a specific probability to any outcome.

We’ve already learned that it can be surprisingly difficult to formulate questions broad enough to be interesting, detailed enough to be valuable, and precise enough to allow referees to clearly see the outcome and award points.

It reminds me of Deep Thought, the city-size “stupendous super computer” of Douglas Adams’s Hitchhiker’s Guide to the Galaxy. It cogitated for 7.5 million years to answer the Ultimate Question of Life, the Universe, and Everything…and finally, somewhat diffidently, coughed up the answer: “Forty-two.”

The answer, Deep Thought assured its dismayed hereditary programmer-priests, is correct. The problem was that the question was poorly defined. They begged the machine to give them the right question.

“Deep Thought pondered for a moment. ‘Tricky,’ he said.”


Join us at SciCast. Registration is free but is required to set up a trading account. Select “IEEE Spectrum” under “Topics” in the left-hand menu. And let the world know what you think.

About the Author

Douglas McCormick, principal of Runestone Associates, is a science and technology writer and editor, IEEE Spectrum’s webinar host, and its test and measurement blogger.

This article is for IEEE members only. Join IEEE to access our full archive.

Join the world’s largest professional organization devoted to engineering and applied sciences and get access to all of Spectrum’s articles, podcasts, and special reports. Learn more →

If you're already an IEEE member, please sign in to continue reading.

Membership includes:

  • Get unlimited access to IEEE Spectrum content
  • Follow your favorite topics to create a personalized feed of IEEE Spectrum content
  • Save Spectrum articles to read later
  • Network with other technology professionals
  • Establish a professional profile
  • Create a group to share and collaborate on projects
  • Discover IEEE events and activities
  • Join and participate in discussions