It's always interesting to look at old predictions– there was that Newsweek article that predicted the Internet was never going to be a big deal [1] (amusing be

It's always interesting to look at old predictions– there was that Newsweek article that predicted the Internet was never going to be a big deal [1] (amusing because now Newsweek is online-only), and there were all the negative comments on the launch of the iPhone. [2] Also I believe Drew Houston's "Show HN: Dropbox" thread has a famous "Why would anybody need this, I can do the same with <complicated procedure>" [3]. It's tempting to think, "Ah, people! So terrible at predicting things." I think it's interesting to think about why that is. The main problem, I think, isn't that people make wrong predictions altogether. It's that it's very hard to see how things will change and evolve over time, and how the ecosystem will change with it. The "One Laptop Per Child" idea [4] sounds a little dated and silly now. I suppose if we just remember that progress is continuous rather than discrete, and that a lot of seeming limitations can be overcome with currently-unlikely innovations, then a lot of predictions will be forced to be a lot more precise. Perhaps the Newsweek prediction might've been amended to, "In its present form, the Internet is unlikely to change the world." The problem with predictions is– things rarely stay in their present forms, and the world around them rarely stays the same, either. I think pg addresses this in his most recent essay, "What Microsoft is this the Altair Basic of?" [5] In his words– "they practically all seemed lame at first." So we have to learn to live in a world where our initially valid assessments of a thing might become rapidly invalid because of change. And this is where Nassim Taleb's work about the problem of prediction [6] comes into play. Rather than trying to predict a particular outcome, it makes much more sense to focus on evaluating robustness and antifragility– "How will this thing respond to change? What are the potential upsides, what are the potential downsides? What will kill it? What will give it more utility?" Even if the odds are really low that something might come around, if the payoff is high enough, it might be worth betting on. I think that's the whole point of things like YC. ___ [1] http://www.newsweek.com/clifford-stoll-why-web-wont-be-nirvana-185306 http://www.newsweek.com/clifford-stoll-why-web-wont-be-nirva... [2] https://web.archive.org/web/20070116071424/http://www.engadget.com/2007/01/09/the-apple-iphone/#comments https://web.archive.org/web/20070116071424/http://www.engadg... [3] https://news.ycombinator.com/item?id=8863 https://news.ycombinator.com/item?id=8863 [4] http://en.wikipedia.org/wiki/One_Laptop_per_Child http://en.wikipedia.org/wiki/One_Laptop_per_Child [5] http://paulgraham.com/altair.html http://paulgraham.com/altair.html [6] http://www.amazon.com/Antifragile-Things-That-Disorder-Incerto/dp/0812979680 http://www.amazon.com/Antifragile-Things-That-Disorder-Incer...