Predictions for 2030

Now that it’s Jan 1, 2020, I’m going to make some predictions about what we will see in the next decade. By the year 2030:

• Deep learning will be a standard tool and integrated into workflows of many professions, eg: code completion for programmers, note taking during meetings. Speech recognition will surpass human accuracy. Machine translation will still be inferior to human professionals.

• Open-domain conversational dialogue (aka the Turing Test) will be on par with an average human, using a combination of deep learning and some new technique not available today. It will be regarded as more of a “trick” than strong AI; the bar for true AGI will be shifted higher.

• Driverless cars will be in commercial use in a few limited scenarios. Most cars will have some autonomous features, but full autonomy still not widely deployed.

• S&P 500 index (a measure of the US economy, currently at 3230) will double to between 6000-7000. Bitcoin will still exist but its price will fall under 1000 USD (currently ~7000 USD).

• Real estate prices in Toronto will either have a sharp fall or flatten out; overall increase in 2020-2030 period will not exceed inflation.

• All western nations will have implemented some kind of carbon tax as political pressure increases from young people; no serious politician will suggest removing carbon tax.

• About half of my Waterloo cohort will be married, but majority will not have any kids, at the age of 35.

• China will overtake USA as world’s biggest economy, but growth will slow down, and PPP per capita will still be well below USA.