Hacker News Reader

Comment by chubot | Hacker News Reader

Really? I wouldn't be very surprised at all. I'm sure the guy has serious programming chops, and the goal is sufficiently vague. He's not literally going to build Jarvis in a year, but there are lots of concrete things you can build to automate your home that are Jarvis-like.

He says "simple", so I would say being able to understand and execute "turn off the lights in the second floor bathroom" probably counts. I'm sure he can do something like that in a year (even though there are programmers who can't).

I think a lot of programmers are under the impression that he is this idiot who knows a little PHP and got lucky building a website. But I'm sure he is probably better than the vast majority of working programmers out there, on account of building the initial versions of Facebook (and I say this as a working programmer). People just don't want to believe that because he seems douchey in many ways.

View on HN

Prior to Facebook, among other things, Zuckerberg made a music player which learned your tastes using machine learning to produce better matches. Pretty sure that even though he hasn't coded in awhile, his coding abilities just need to be reawoken.

Not to mention that he had built several other things prior to starting Facebook.

There's also a lot of libs/services he can use

And if he needs help in understanding or implementing something, he can always ask for some help from someone @ FB...

This. I think most of the work that goes into building a successful large-scale project is making the design choices that minimize required effort.

Everything after that is a matter of typing correctly and adjusting design choices to unexpected constraints.

Having the knowledge of FB's engineering at his disposal will certainly make things easier.

Facebook acquired https://wit.ai/, which allows you to convert sentences (in speech/text) like "turn off the lights in the second floor bathroom" to structured intents. I'm sure he can hook that up to an arduino or Raspberry Pi with a relay in a couple of days.

The guys of wit.ai are now in the Facebook M team, I've read somewhere last month (can't remember the source though)

I bet against you. In the real world in 2016 even if you are a genius you can't do it alone except in very rare cases. You can impress others with your individual skills but you can't push the boundaries.

If he can achieve his goal will mean that a lot of people could do that and there is nothing special in that project. You can argue that he has more access to data than you and me but several people working at FB have the same access as him. His simple AI will be... yes, simple.

> If he can achieve his goal will mean that a lot of people could do that and there is nothing special in that project.

I think that's kind of the point of his personal challenges. He's trying to set an example of being intellectually curious, not show off that he's a genius. Wasn't one of his previous projects for the year to learn Chinese? A difficult challenge, but not a superhuman one. Plenty of other people are capable of it.

There are things to dislike about Zuckerberg, but I don't think these challenges are one of them. They're pretty admirable imo.

Reply on news.ycombinator.com