If you think that smartphones/mobile have peaked… this article makes a strong claim otherwise. The summary: tech cycles go in gestation and growth periods. Since around 2006 or so we have been in a gestation period. Smartphones are omnipresent, the tech that is used for smartphones has made everything smaller and cheaper, and the ecosystem is ready to go…
“The business plans of the next 10,000 startups are easy to forecast: Take X and add AI. This is a big deal, and now it’s here.” — Kevin Kelly
Time to dust off those coding chops I guess. There are so many venues in which some basic Artificial Intelligence would be useful:
- Watches and other wearables. We aren’t going to type on those things.
- Home automation. Who wants to find their smartphone to turn on a light? Amazon’s Echo is already doing this to some extent (seriously go try it!) and this will only improve and get better with more competition.
- Scheduling for business and travel. We schedule meetings basically the same way in 2016 as we did when I worked on Schedule+ in 1996. Companies like http://x.ai are working on this but it’s still early. Scheduling travel is in some ways *worse* than when you called a travel agent. How much do you think giants like Expedia or SABRE are looking at this (scratch that: since SABRE is an airline consortium they will no doubt be shocked with the development of new stuff).
The role of open source
One area I didn’t see in this article is the role of open source. In the last few gestation–>growth periods one of the hallmarks of the transition was open source would get good enough to power growth i.e. the frameworks would be good enough. Coding a web app used to be hard. Now you can grab a bunch of open source tech, some of it hosted by very viable companies, and turn out a project in a few days. What are the key open source technologies we will need to take mobile to the growth stage?
One key item: data. The article notes that the more data you collect (e.g. people talking into their phones a la Siri/Google Now/Cortana) the better AI you can do. What this means is that the smaller teams, the real innovators, will either need to get into a niche or the open source community will need to find a way to share this data. Imagine an open data mart that little teams could participate in so they didn’t get overwhelmed by the massive data models from Apple, Google, etc.
This worked well for Wikipedia, web platforms, and code: what is the data mart for AI?