Intelligent Agents are coming
Intelligent Agents are coming and you don't have to be afraid of them. You should try to understand them.
For two years, disserting about artificial intelligence has become invitable to shine during meet ups, investor pitches or in executive committee. However, those who talk about it and really grasping what is at stake are even scarser than those would really understanding how Google algorithm actually works.
In the same time, how can you avoid talking about it as there's a strong social pressure that leads us to do so (You're not working on a chatbot prototype ? Do you want to become obsolete ? Haven't you realized that Google became an AI first company 18 months ago ?). How resisting the envy of believing that the era of artificial intelligences has finally come : some Silicon Valley Gourous argument that AI is the worst threat on humanity that ever existed, expert goverment reports claim that 10 %, 20, 30 % of jobs would be in jeopardy in the next 15 years, GAFAM companies (Google, Apple, Facebook, Amazon, Microsoft) and the greatest investors bet billions on AI.
So is AI a balloon that could eventually deflate just the way Yahoo, Google Glasses or Uberization did ? Or, is it a deep-rooted trend, just as social networks or marketplaces ? And, is AI really threatening us ?
Reality is more subtle : AI currently is a great experimental worksite where thousands of ingeneers work and wehre genuinely intelligent applications are scarse, but where some specific applications could boost productivity for some activities or enable some verticals to rapidly change. In other words, the « 2001 the odyssey » Hall 900 or the « Terminator », won't come in 2020, but maybe as soon as 2018, automatons or almost intelligent agents will work out some very simple tasks with or in place of humans. Maybe n 2015, they will have profilerated and maybe in 2030, they will begin to merge and simulate 2 % or 3 % of human intelligence. This may seems a very litte figure, but keep in mind that 2030 in only 12 years from now and artificial intelligence enthousiasts have been waiting for this miracle for the last 50 years.
For this to happen, at least 4 conditions are to be met : enough Steve Jobs in the AI field, enough capital, profitable business models and a very small detail : that human accept to offert some space to bots, automatons and other robotic intelligences. Failures in most technological revolutions announced by visionnaries are to be explained by the opposition of users, who have the very bad habit not to rush into any innovation that offer that tech companies produce. They refused to buy the iphone grandfather that Apple launched in the middle of the 1990s, got stuck in front of Google Glasses but loathed ridiculous augmented reality ears in Snapchat. In the same time, this is all the less surprising when you remember that they massively adopted « low value » services such as text messages. Technological revolutions would actually not happen if consumers don't give their go for them. Some even whispers behing the scenes, that millenials, so often presented as techno enthousiasts, would be reluctant to the majority of new revolutionnary apps and that most start ups would fail....
Today, anyway, artificial intelligence is just currently turning the the status of keyword that ease start ups fund raising or sale of consulting services from specialized agencies, to the status of actually working applications for humans. Besides the GAFAMS companies who use very specific artificial intelligence technologies to predict future users expectations to push their targeted ads or always more addictive contents so that they spend even more time in front of their screens, artificial intelligence success stories in the real world can be counted on the fingers of one hand.