Placeholder Image

字幕表 動画を再生する

  • Simon: So I'm here with Jeff Hawkins, this is the guy that invented the Palm Pilot...we've

  • seen Mark Shuttleworth talk about his edge phone. This is the guy that created the market

  • Mark's trying to sell into.

  • Jeff: We actually created one of the first smart phones, the Trio.

  • Simon: ... ... you don't do palm pilots any more - tell me about your true passion.

  • Jeff: my true passion, and has been for over 30 years, is neuroscience. Understanding how

  • the human brain works, (the neocortex) and building machines that work on those same

  • principles. So, the whole period of palm and handspring was like a sideshow... I couldn't

  • get the gig I wanted in neuroscience, so I was building mobile computing. I ... loved

  • it, I enjoyed it I was totally excited about it, but my real passion was brains, how they

  • work and building machines that work on the same principles.

  • Simon: So now you've been working most recently on software that can predict the future?

  • Jeff: That's right, we've actually been working on modelling, figuring out how parts of the

  • neocortex works. Which is the big (rich?) region on top of your head and we figured

  • out first what it does and then how it does it (lay yourselves there?) but we've been

  • applying it to problems where you can take streams of data, model the data and then predict

  • the future. So we actually have a product called Grok which takes data from windmills,

  • energy meters and other machines and it can predict future values it can detect when anomalies

  • are occurring and things like that. So we have a sort of business side to what we're

  • doing, but we're here at oscon to talk about the new open source project where we take

  • these learning algorithms, which are essentially models of the neocortex, a slice of the neocortex

  • and putting them in an open source project.

  • Simon: So that new open source project, you're launching it?

  • Jeff: pretty much, it's been up for a couple of months now. its called "nupic" the numenta

  • platform for intelligent computing. its the same code tree that we use in our products,

  • so you can go and see what we're releasing every day we have an active community already,

  • we've had our first hackathon, its only been going for a couple of months, but its been

  • going pretty well so far, we're pretty excited about it.

  • Simon: So what would I use that for as an open source hacker?

  • Jeff: There are a number of reasons why you might be interested in this, people are interested

  • in applying it to new problems. We're applying it to machine generated data but theres lots

  • of other applications where people might apply it to. You might apply it, in the same vein

  • as we're using in Grok you might apply it to financial type of data or you might apply

  • it to new sources of data which we don't look at, we're focused on certain things. There's

  • a lot of people interested in taking it and building more complex systems: robotics, vision,

  • music, things like that. and this requires extending the algorithms. its a memory system,

  • so making them bigger, putting them in a hierarchy and so on. So theres people who want to apply

  • it to existing products, like, i want to predict something, people who want to build new types

  • of products, people who are interested in language, semantics and so on. then there's

  • people who are interested in doing pure research, they're doing mathematical analysis of these

  • algorithms, so its kind of broad you know, we're talking about the beginning of building

  • brains in software and hardware and that's a big big field, it's going to be huge and

  • it's just getting started. And I should mention we have a number of people interested in doing

  • hardware implementations of these algorithms there are some big companies, ibm, cgate,

  • some others who actually have programs on the way, because they're pretty excited about

  • this stuff.

  • Simon: So I'd sum it up as "using brain science to work out how to handle big data".

  • Jeff: It's one way of looking at it. I'm a neuroscientist so I always want to talk about

  • the neuroscience, but from a hacker or coding point of view yes, today what you can do with

  • it is yes you can stream fast data to it and it builds models of the data in an online

  • fashion meaning every record that comes in its updating the model, it makes predictions

  • and can detect anomalies. I'll give you an example, a simple example we actually did

  • and used it for is people are interested in predicting energy usage, so a building consumes

  • energy throughout the day, its up and down depending on what the buildings doing or what's

  • going on, if you can predict what the energy consumption will be 4 hours from now or 24

  • hours from now its sometimes advantageous. You can pre-cool the building, you can do

  • a thing called demand response where you basically buy energy at different prices. so thats the

  • kind of thing we do with our product Grock today, it works very well at that. Constantly

  • learning and if the patterns in the world change it adapts to it automatically.

  • Simon: So what was it that made you give up palm pilots (which were awesome) and get into

  • neuroscience instead?

  • Jeff: Well I think that brain science, understanding the brain and how it works and building intelligent

  • machines is actually a bigger societal impact long term than mobile computing. much much

  • larger. You know, mobile computing, absolutely, everyone in the world is going to have a computer

  • in their pocket, it's a process for democratisation and education, so thats all great, but people

  • don't realise yet how big intelligent machines are going to be. It's sort of like starting

  • the whole computing industry all over again. So in my talk I mentioned that we're like

  • the 1950's in computing. 1950s in computing was when they were just starting to build

  • computers, they were just starting to be useful, but we had decades of advances still to go.

  • We're starting to build intelligent machines that work on the principles of the brain,

  • we're just getting started and it's going to be decades, but where its going, its just

  • gonna be unbelievable, we're going to be able to make machines that are million times faster

  • at thinking than we are we're going to be able to make machines that have much more

  • memory than we do, we're going to be able to make machines that can sense things that

  • we can't sense and so its hard to know where its going to go, just as in the 1950s it was

  • hard to know where the computer was going to go. but intellient machines, machines that

  • learn in the way that brains do is just going ot have an amazing impact on society, on the

  • earth and humanity.

  • Simon: How do you think being an open source project is going to contribute towards achieving

  • that vision for you?

  • Jeff: First of all, my goal has always been to make this happen sooner, to be a catalyst

  • for this. So anything i can do to spread ideas is a good thing, I'm not in this at this point

  • trying to make a lot of money off of this. I'm in this because I think its cool, its

  • fun, its good to do its important. I waited, even though we made this technical, scientific

  • discovery four years ago and we published it, i waited until we had real demand before

  • we made an open source project, I wanted people to come to me, and they did, people come to

  • us and said can you give us the source code to this, its really cool, we want to work

  • on this, i want to use my phd thesis on this, we want to embed it so when people asked us

  • we said great and of course this was my goal from the beginning. You can't help but put

  • these ideas out there. Putting the code out there, showing how this stuff works. some

  • number of people are going to pick it up, they're going to go, "this is great, I get

  • it", they're going to invest in it. It's not something anyone can own, it's not one thing.

  • it's like saying, "could the computing industry be closed?" no, it couldn't, it had to have

  • lots of competitors, lots of ideas, and this is like that, this is not something that any

  • person or company can own, so its got to be out there.

  • Simon: its been fantastic to talk to you and great to see you here at OSCON, I wish you

  • every success with the project you're working on, thank you very much for talking to us.

  • Jeff: thankyou, It's Nupic and numenta.org is the url

Simon: So I'm here with Jeff Hawkins, this is the guy that invented the Palm Pilot...we've

字幕と単語

動画の操作 ここで「動画」の調整と「字幕」の表示を設定することができます

A2 初級

ジェフ・ホーキンス - ビッグデータで未来を予測する (Jeff Hawkins - Predicting The Future With Big Data)

  • 318 18
    Yang Allen に公開 2021 年 01 月 14 日
動画の中の単語