Articles

Basketball With Watson


Whoa. You guys have seen this before. It’s called a Sphero. Works with this cool app, you drive it around. You can go off jumps, play games, but believe
it or not, this thing is more sophisticated than it looks and today we’re going to test
that out. I have teamed up with IBM for this video. We’re going to be using Watson to take this
Sphero to the next level. Now many of you know I am a Toronto Raptors
fan and the Raptors actually use this Watson tech to do player evaluations and make decisions
based on data, so we’re going to carry forward with that basketball theme and we’re going
to play Watson basketball with the Sphero. The table, as you can tell, is now a basketball
court. Thankfully, Jack is somewhat artistic here. This little Sphero is about to become a basketball
player. How exciting is that? Through the power of Watson, the Watson API,
we’ve got an application that means he’s going to interpret my commands as if I’m
his coach and then react like a player. Maybe even experience what victory feels like. Now obviously this is not what the Toronto
Raptors are doing. Actually, I don’t know. Maybe this is exactly how they make their
decisions. Probably not, no. they’re doing far more sophisticated stuff
crunching data. They’re using Watson APIs in order to determine
sentiments surrounding a particular player to look at stats and crunch it all down, so
they can make better, more informed decisions. The various instructions result in different
behavior for the Sphero. For example, certain instructions will result
in him changing color whereas other ones will result in certain actions. Let’s get the game started with, of course,
an important component in basketball – the tipoff. Watch this. Tip off. I got this. Let’s get the possession early on. You saw him turn green. You heard the response. Interpreted that language. Game’s now started. I need to tell him to make a move on the fast
break. Let’s try this out. Fast break. On it. They won’t know what hit them. Look at that. Moves right over to the offensive zone, okay. He’s down over here, but what’s the next
move? What about defense? You’ve got to come back every so often. You’re trying to be a two-way player. Alright, Sphero, play some defense, please. Yes, Coach. Tracking back now. Oh, you see the way that happened? Right to the rim, he was tracking the whole
time. You’ve got to keep the shoulders in front. That’s a tip for you kids out there. Eyes on the chest, never mind the ball. You don’t want to get crossed up. Alright, Sphero, the other team sent the ball
out of bounds. Please head down there and inbound it for
us. Turnover. Time to go on the attack. Whoo-hoo, he jumped straight out of bounds. The cool thing here is it’s actually pretty
sophisticated, so even if you don’t necessarily phrase things in an exact way, there are key
words that are picked up as reference points. The Sphero can still interpret what action
it is that you’re looking for. Something I want to mention at this point
as well is that this app is browser-based, so theoretically it could run on a laptop,
tablet, an iPad, an iPhone. I told you how in professional cases like
the Toronto Raptors, there could be some sort of sentiment analysis tracking social media
in order to interpret what’s being said about a particular player. Okay, Sphero, job well done. You won the game. It’s time to celebrate the victory. Wahoo. Knew we would win with you as our coach. That’s right, Coach Lew. Maybe I’ve got a future in this. Look at him dancing around, turning red and
if the Watson API is picking that up as sentiment off social media, Sphero’s going to celebrate,
so please send your tweets, #victory. What you just witnessed was three separate
Watson APIs in action. First, speech-to-text, that’s me speaking
to the Sphero. A conversation API allows the application
to understand what I’m saying and then respond. The text-to-speech API actually converts text
from the app into speech, so then I can hear what the app has to say rather than reading
the screen. Truthfully, we’re just scratching the surface. This stuff is used at all kinds of different
companies. It’s even used for music recommendations. If you want to check it out for yourself,
there’s a link in the description. Go find out what you can build.

Leave a Reply

Your email address will not be published. Required fields are marked *