Video Of The Week: Larry Page at TED
Tyrone suggested I post this today and two minutes in, I was sure it was the right call.
So here is Charlie Rose’s interview of Larry Page at TED last week.
Tyrone suggested I post this today and two minutes in, I was sure it was the right call.
So here is Charlie Rose’s interview of Larry Page at TED last week.
Comments (Archived):
Geez, the make-up department at TED has outdone the one at LeWeb.
His humbleness admitting there is so much we still don’t know is impressive.”Computing is a mess and Google is helping?” Hmm. That’s what Google is exploiting!Other points that caught my attention:”Throwing intelligence back at us.” – Like”Searching for a better life for me and my neighbours.” Like”Universal anonymous sharing of all medical records.” – Like”Separating bikes from traffic.” – Like, but on a wire?”We need revolutionary change, not incremental change.” Like”What many companies do wrong is- they miss the future.” – LikeQuestion for Larry: If we have a mesh of balloons around the world, won’t they collide with planes?
@wmoug:disqus as per http://googleblog.blogspot….It’s very early days, but we’ve built a system that uses balloons, carried by the wind at altitudes twice as high as commercial planes, to beam Internet access to the ground at speeds similar to today’s 3G networks or faster.
Interesting. Thanks for the link. It’s definitely experimental.
Their CEOs can pull this off. he’s a founder/entrepreneur with a $8 billion exit under his belt, medical doctor and a real innovator. He knows how to get things done
Maybe we will be biking between the balloons. Catch a tailwind and cycle @ 45 mph @ 1000m?> If we have a mesh of balloons around the world, won’t they collide with planes?
Nice summary. Would be nice if there was a service that provided transcripts of these videos.
There are several it seems:https://drive.google.com/fi…
I think the 10x revolutionary change point is important. The thing that really stuck out to me about Larry is the fact that he sees all these problems as opportunities for people and companies to tackle. I think too many companies and entrepreneurs are just roaming the harbor when there is a giant sea to explore.
As inspiring as Google is, I have to think that LP and SG must be frustrated that they are so dependent on ad revenue. They still havn’t cracked the ad ceiling and developed a product with mass appeal that (hundreds of millions of) people will pay for.
I’m not sure they’re frustrated by that, as they could always charge for Android but choose not to (as far as I’m aware). Lots of organizations will pay the fee for Google Apps for their domain. We used it on the last campaign I was on and now at my current job. But you are right, the ad revenue is so large it eclipses these victories.
A system for sharing medical records is being done by Nanthealth
Wow. I just checked them out. Pretty ambitious company. Thanks for pointing out.
I didn’t know about his voice before this. Interesting how Sergey says he’s a better CEO now because he chooses his words more carefully.
You meant, Larry… no?
“Page adds that his ability to exercise strenuously has also been “somewhat reduced” because vocal cord nerve issues can affect breathing. “That said, my friends still think I have way more stamina than them when we go kitesurfing,” he writes. “And Sergey says I’m probably a better CEO because I choose my words more carefully. So surprisingly, overall I am feeling very lucky.”(http://www.businessweek.com…
Thanks @fredDef listen 27 min in to this – http://youtu.be/MBItc_QAUUM…Steve Jurvetson: I was at a lunch at a Google event, and out of the blue, with no expectation this would be a topic Larry Page turned to me, knowing a little bit about our connection, said, “You know, how much money do I have in you?” He mentioned a number, and I thought that was cute, um, that he was trying to recall that. He goes “You know, if I were to get hit by a bus today, I should leave all of it to Elon Musk.”Elon: Really? Steve Jurvetson: Yeah.Elon: He said that? Steve Jurvetson: Yeah. And, and so I’m like, umm, paper & pen, we’ve…(Elon: AhaHaHAHA!) please jot this down! Yes! Elon: I know that actually… he’s a good friend of mine. You know, I met Larry before he got venture funding. Steve Jurvetson: Really?Elon: Yeah. Steve Jurvetson: Wow. Well he’s a remarkable guy, obviously, also an underachiever. And, umm, you know, has a company that wants to do good in the world.Elon: Yeah.Steve Jurvetson: And I think he looks at you with a bit of envy. Because what he then proceeded to say was “You know, I could.. give my money to a non-profit and a lot less will get done than a corporation that’s pursuing things that are directly aligned with things I care about.”Like, getting off of oil and colonizing other planets. He believes in those missions. And thinks, that a corporation, endowed with a right to do that as its business purpose is the best vehicle out there. And he wishes he could do more of that in his own life. He compared, poignantly I think, to some other software companies in the pacific northwest who might have executives do evil for the first part of the career and then do good for the second half, and then the sad story of others who never get to the second half of their life, like Steve Jobs. Steve Jurvetson: Ummm, uhh, not in a joking way, I mean, seriously, and it was a very deep moment.And here is something more light hearted and funny for today.https://www.youtube.com/wat…
i hope he can find a cure for his voice problem. it must be so frustrating.
Thanks for the vid. Google is very much on the forefront today and it has been interesting looking Google’s different press releases moving into this year showing the true game floor they plan to be top player on.I did a presentation late January that got mixed reviews where I took the focus of AI over to Google (did include IBM) with Deep Learning, Cats, NG bringing down the cost of Cats and so on. I’m going to send this vid to the teachers for them to show when there is free time.Fred, last year you stated your position regarding Google. I’m curious if you had a preview of what was coming as we moved through the holidays over into this year.
The transportation piece struck me. Man doesn’t make great strides forward without better communication and transportation. We haven’t had disruptive transportation in a while. Everything iterative. Driverless cars are disruptive……
I’m thinking that advances in drones [1] (60 minutes had a piece on a medivac drone prototype) will lead to a situation where the dreams of people flying their own planes in the early part of last century that never panned out will actually come true.One of the issues with everyone owning an airplane (or actually helicopter since they don’t require runways) is not only the cost (which would for sure come down with scale) but the safety and human error factor. It’s more or less a non starter for mass adoption.But I think with what I am seeing with drones it is entirely possible that transportation will move in the direction of air travel. If we can improve on the safety factor (and we definitely can with the advances that I have seen just in the last 5 years) then the product becomes something that the masses can and will buy.I guess that’s actually pretty obvious at this point, right?[1] Missy Cummings, who runs the drone research labs at MIT and Duke, thinks that drones are no more invasive than CCTV cameras, Facebook and cellphones.At a drone show, Cummings shows us the most exciting thing: a medevac (medical evacuation) helicopter that can navigate itself, load a passenger and bring it to another destination.
yup. agree on drones
Just a small technical note: air travel is incredibly energy intensive, particularly for helicraft since they don’t receive the lift benefits of wings. Also, helicopter engines are sized for take-off (max lift required), and are throttled back massively during cruise, which makes them inefficient. You may be able to solve the efficiency problem with electronic motors (like drones have), though that would require a battery breakthrough as well.Driverless cars excite me because when you decouple the driver from the car, you can decouple ownership from the driver as well. Combine that with the massive energy storage capacity of a Tesla Model S type electric car, and you have an on-demand fleet of deployable vehicles, that can also be used to provide Vehicle-to-grid energy storage. This is an incredibly disruptive combination of features: transportation, energy, sharing economy
Nice video wish I had the time to watch the entire thing.I just skipped around a bit and found this interesting:Page: I look at medical records and I say wouldn’t it be amazing if everyones medical record were available anonymously to research doctors and when someone accesses your medical record a research doctor – they could see – (um) you could see which doctor accessed it and why you could maybe learn about what conditions you have I think if we just did that we’d save 100,000 lives this year [1]Rose: “absolutely”.Rose wants to move on but Page persists. His point (which you couldn’t really get from what he said, as he doesn’t communicate that well, at least in the few parts that I saw) is that when he made his condition public all this info came out of the woodwork to help him.I’m not sure I understand the benefit (someone please explain) of being able to see which research doctors took a look at your medical records. If anything that would seem to be a good way to prevent the researchers from looking for fear of being inundated by people wanting to contact them and know why and/or ask for help.[1] The figure of 100k lives is pulled totally out of the air and is obviously arbitrary.
i agree it’s a data point which could be mis-interpreted. it could mean a variety of things. as long as the patient knows that, it’s ok.
Same concept as when you see someone is pinging for your credit score.
I don’t think that is what he means. I interpret what he is saying (and in fact he says “anonymously”) as meaning that a research doctor without prior authorization could access your records without knowing it was “you”. But you would know the doctor that accessed your records.Right now you don’t know which companies access your credit score (other than they will send you credit card offers). And when someone accesses your credit report you have generally given them permission (the local bank when applying for a loan).Most important point though is this. If the name of a person is associated with viewing your information it will almost certainly result in traffic for that person “Dr. Hayes” or the institution “Hayes Center for Research”. This is quite different than even a similar situation where a large unapproachable company has access to that same information “National Institute of Health”.
I took it in that if a doctor who perhaps has noticed a pattern or whatever could then search for that pattern amongst everyone’s records, and then perhaps have a way to contact them. This does open up a problem though. There can be value in this level of friction because it makes things move slowly and will require a doctor or researcher to more and more prove or show to others the legitimacy of what they’ve found before the solutions would more broadly be applied.
Cross tabulation and more such brute forceexploratory data analysis is a poor way tofind real causes or good causal models.E.g., imagine getting the data on the observedpositions of the planets in the night sky and,then, from that, via just data analysis, discovering Newton’s secondlaw of motion and his law of gravity. Or tryto imagine how the heck simplistic data analysis could come up with Einstein’s general relativity — partly he was guessing,early on passed some good tests, butsince then his work has proved fully validover huge ranges, far beyond any datahe had.
If you an replicate the true/false and question-experiment layers of the brain I am sure you could have a system that could discover the questions to be asked – it just happens that some minds and brains have the base function and patterns to evolve in a way that their experiences lead to answering specific strains or contexts of information. Theory wise I have a model I’ve been evolving on how to do this, I won’t have time for at least 10 years though likely to build it – and I have another layer to a mind-body theory I have already that I want to create a computational model for. It’d actually be great and helpful if I developed them at the same time as they could become an equation.
If you want and can, then go for it. No doubt someone some day will figure it out — clearly the it here is real AI — and successfully make math and, then, software out of it.My view is, it’s darned hard. What I’ve been working on is much easier, hard enough to give me a good barrier to entry but easy enough that I have all the core software in good shape. I’m just trying to make money; what I have is, like some of math, quite general and, thus, if successful stands to make a lot of money, but that was not a requirement when I started although “welcome just the same”.For starting on AI, one approach is to let a natural language be the main means of ‘knowledge representation’. Another approach is to start with something simple, maybe even a worm and work up to a frog, bird, etc. Another approach is to start at ‘birth’ and track how animals grow up and use that to see how the software has to ‘learn’. Still I think it’s too hard.
I’ve seen some pieces that are needed already have been created, it’s a matter of tying it all together properly though. I think I’d have a good chance of figuring out more of it if I ever have the opportunity/time to.
Signal to Noise Ratio: What is clear as day to me is that Page never looked at a medical record. Medical records as they are constructed now are mostly noise.
Its the same idea that powers PageRank. Not particularly insightful.
Not related, just since it’s weekend-http://www.youtube.com/watc…
11
Thanks for the help, @tyronerubin:disqus.
I have to disagree with him on companies being viewed as evil. I think there is a segment that generally views them as evil but in many circles companies are revered. People love their automobiles, favorite foods, and Apple has a cult-like following.
I’m not sure its a dichotomy.I love and enjoy technologies, and capitalism can be a fine way of organising resources to achieve the goals that deliver those technologies.But capitalism is normally assumed to function in “production” rather than R&D mode leaving a mismatch when the rate of change is high.As a consumer I pay not for what the organisation will do with my money, but to reward what it has done with previous investors money. If this is taken as approval for their current aspirations the assumption is incorrect.When US citizens paid taxes to sponsor NSA espionage, was that what they “thought” they were buying? Clearly not.There is a problem of governance if organisational leadership assumes that current revenue streams map to approval for current activity where there is a lack of transparency or a deferred feedback cycle.
It is impressive to see Larry grows into his role as Google CEO. I think he is doing a better job in steering Google towards the future than Eric Schmidt. Full-disclosure: I was a former Googler under Eric.
His ambition is so inspiring – we all know the george bernard shaw quote about asking “why not” – the frequency with which he does that and changes thing is just amazing. A wicked smart learning machine..Thanks for sharing this – Tyrone and Fred!
For the computers to really understand us and our information, it would have to go deeper than being able to contextualise on geo-location and time positioning which is what Google Now, Deep Mind and most Internet of Things sensors do.It would need us to have a more coherent view on how our brains work and the diversity of that intelligence.At the moment, the mathematical supposition that underpins our brains and the machines is that it’s logical, rational, probabilistic, geo-spatial and other characteristics that would enable it to pass an IQ test on a conformed basis (overlooking cultural, gender and socio-economic factors).So……a nextgen machine would need to have a re-imagined framework for the brain and also apply a new form of mathematics that is not about logic, rational numbers and probability.
Evolution takes time. Just the fact the machine is now able to learn how to play the game is a step forward.In the future, if we have the machine within the machine (conscience) as they have apparently found within the human brain, where it asks basically is it this/that leading to the “why” regarding this/that. That is the road to true autonomy.You just can’t get there on the flip of a coin, so be thankful that you will see something in your lifetime that is so incredible.
Thanks, Dave. It’s a good thing I put “WHY?” at the core of the intelligent system I’m building.No, we can’t get there with flips of coins and training the machines to go down binomial (0, 1) outcome paths or cellular automata transmitting over x-y-z pixel positions.
We had the special on Temple Grandin shown on Discovery or SciFi regarding Innovators saved for a long time and watched it last night. It was fascinating watching her go through imaging determining the amount of connected activity between the front and back of brain. It was even brought up how her line of thinking is similar to deep machine learning regarding the use of images.
what is ‘TED’, and the ‘x’ that sometimes follows?
TED stands for “Technology, Entertainment, Design” and is a conference in California (apparently now in Canada) where people gather and talk about ideas. When they put the talks on their website some years ago it “blew up” I think because there was a hungry audience of nerds spread across the US that would love to learn a tidbit about something during their lunch break. TEDx is a variant where the organizers allow people to hold their own versions at their companies or universities. There has been much discussion about it with critics, etc., but I think the main point is to distill an interesting idea or experience into a digestible story.