MyLifeBits

Project Infobox
Self researcher(s)	Gordon Bell
Related tools	SenseCam, pedometer
Related topics	Cognition, Social life and social media
Builds on project(s)
Has inspired	Projects (0)

Show and Tell Talk Infobox
Featured image
Date	2010/08/11
Event name	Bay Area Meetup
Slides

This content was automatically imported. See here how to improve it if any information is missing or out outdated.

MyLifeBits is a Show & Tell talk by Gordon Bell that has been imported from the Quantified Self Show & Tell library.The talk was given on 2010/08/11 and is about Cognition, and Social life and social media.

Description

A description of this project as introduced by Quantified Self follows:

Gordon Bell, legendary lifelogger at Microsoft and author of Total Recall, gave a candid, engaging talk on his MyLifeBits project. He showed pictures from his SenseCam that takes a picture every 20 seconds while he's wearing it. He generates 1 GB of data per month, including a screenshot of every website he visits and the hundreds of pages of health records from his multiple bypass surgeries. Gordon considers his data to be a surrogate memory - he doesn't look at it unless he needs to remember something specific, and he enjoys making films from the pictures.

Video and transcript

https://vimeo.com/14282561

A transcript of this talk is below:

I’m a principle researcher at Microsoft and I think my tools are just computers. Anyway, this is a project and I’m going to talk about the quantitative and self-data what I’ve been kind of working on for the last decade, which was a project called MyLifeBits which was the issue of trying to realize and that was trying to put everything basically on your hard drive.

It started out with a scanning process, and if you look at some of the stuff in there, there’s everything from the back strategy. I used to head engineering at Digital Equipment Corporation. It’s got everything and I tried to get rid of everything and I’m not going to talk about that aspect of quantitative, but this is some of the stuff that’s in there and some of the content that’s in there. Towards the end we’ll get into some of the more quantitative stuff and I’m just going to give you things that are basically quantitative as opposed to these sorts of images which there are about 100,000 in my hard drive. I worked with Jim Bray, and so this is a database thing, and Jim as you may know was the inventor of SQL. And the issue was can we use SQL to put our life in there. And the upshot was by 2005 I believe that SQL or the system that we were building was basically a transaction processing system for life. So I tend to think of the whole thing is that you want to capture everything, and we had lots of experiments to try different kinds of everything in a way that basically everything is in this database. Everything, you know from mail and social stuff, and I regard my hard drive as my ultimate ground truth that’s e-memory, and I regard this as a URL to my e-memory and metadata. So my physical memory is one that controls are the accessor for my e-memory. All of this, I was looking at this as in terms of a context one has a bunch of different lives here. We’ve got this memory that accesses this other memory, and then we’ve got a personal state of an individual which is location, heartrate, blood pressure, stress, GSR – all of the things you want plus the physical stuff. I’ve got a large heart here because I had a cardiac and I’m into this because of the health aspect so I’ve got a lot of health dat. What’s in there is I tried to sort of look at the kinds of things in there, and I divide the world into work, I divide the world into learning things that you need. And health is a biggie and then just plain old life, everything from chatbox to memorabilia to all the media that you have. What I was trying to show was how much I was doing, how much I had done, like telephone; I did a few telephone calls. I’m an engineer, so fundamentally I look at it from I’m going to try it and does it have any value to me. I talked to Kevin at one point and he said watch you’re not recording everything and you should be. I was like yes I should have done that but I didn’t. Anyway at that point I wasn’t finding it useful and at that point we weren’t able to read it. But I have tried to record everything, and anybody who know of the project knows that this is a sense cam and they would have come up to me and said, you’re not wearing your sense cam and so I preempted that question by wearing my sense cam, and I’ll tell you a little bit about what that sense cam does. I’m going to talk about some bits; some of the bits are health bits. Some of the things that you get from and we were talking this morning and said, well gee you’ve got all these photos, how do you use those in everyday life. And I’d say, the bottom line, the killer app there from my way of thinking is, I’ll get to the point is screen saver. So that thing is part of a refreshing – don’t laugh it is bad. The other app is infact that if you’ve got the time to make movies, Ken Berg style movies or just ambient video and such. So the sense cam is a device that takes pictures every 20 seconds. It has all of this stuff here; light levels orientation, and different color bands. And then it takes pictures every 20 seconds when it thinks it’s not going to have a blurry picture, and it’s successful. It’s not 90% of the time. This is a sort of a full regalia, and this was built by a Cambridge research group at Microsoft. And it’s being used in a number of applications and they all have to do with memory. There’s a company now that’s called Vicon review, which is actually manufacturing and it’s a little company in England and that’s being used to help people with impaired memory. And I can show you some of the movies that I’ve done, but this is one of the movies of just riding around in Cambridge on a bicycle, and this is kind of what it produces. I find that it’s useful for this kind of thing because everybody has a name tag, and so if I take a picture of you and the name tag. I’m not going to ask you to sign a waiver or anything like that. But I find it useful for particular kinds of walks going to trade shows, going to meetings, but more interesting strolls and things like that, where it’s something of using your camera, so you can record your entire day. People use this now and there’s somebody who has used this for five years, and they have 6 million images, and their research and they wouldn’t part with it but I’m not at that point yet. So there is a researcher who has used it for five years, and it’s in Dublin City University. Fabulous work on using it and we can talk about this thing at the end, but fundamentally I think that in a decade I can see that this thing can actually be useful. This was my girlfriend and now my wife and took a sense cam image and won’t show you that of how I was working with it. The pacemaker people get my pacemaker right as I went in for a bypass. What’s in my medical records is scanning everything from 1941 which many of you think there wasn’t a period that old. But it’s got my stuff from 83 in there and my second medical record was about 468 pages. I’ve got a lot of x-rays and CTs and other things, and a lot of files and folders and things like that. All of you probably have your genomic reports, and what I find interesting about that is to find out which you’ve got, and then I get these periodic updates and some were very good it says, I’ve got a high probability of heart attack – good for diction; I’ve had two of those, and two bypasses, and two pacemakers. So anyway it’s pretty accurate there, so look at these other things. And then every year or so they send me an update that said you thought you were in the clear. So as I said I’ve got heart things and this was my first letter on that and my mother had to keep it. This turned out to be useful; this little sketch was my first bypass and handed it to my Stanford doctor a couple of years ago and he said, yes that’s useful and we like to know what’s going on in there before we sort of open you up this time; so that’s why you want to keep stuff. At one point I found out there was a break between a couple of chambers here, so basically they had to put a wire between them, which has a little computer on it. Talk about data, this is every heartbeat here, and here is a particular 18 million heartbeats and you end up with about 1.3 billion heartbeats and some people think that’s a constant for everybody. Anyway, the interesting thing about this is all of this data, this distribution, looking in how this is distributed and what you’re doing they throw all of that data away. Where is that data. This is actually a scan from a floppy and a print out, so as I leave the the unit and the sheath that I scanned and all of this data goes on a floppy somewhere or are on my floppy. This data here could be useful to say, how are you doing now versus six months ago or your last visit. Or what’s happening to this distribution because no one cares until I get sick; nothing is important. Anyway I needed a manual I didn’t think they had it set right, and so I went and got the manual and yeah, sure enough it was a beast to work on so there’s like 16 modes of a pacemaker. So if you want to go out and understand pacemakers I decided I didn’t want to program that, and the company kept changing the interface and all of the data and all of that. Anyway the number you look at on all of this is and that’s my last one and I’m very proud of that and the maximum of 10 years maybe eight and a half estimated. And it’s going to run out the same time I do, so all I’m praying for is this will hold out long enough, and that’s some of the other stuff. Why did all this happen, well after it happened I said well I better start measuring cholesterol and I used to do that and then I measured a bunch of those and then I started taking drugs here. And now my cholesterol doesn’t matter because it’s down low enough and all of that. When I was in the hospital because I got an infection, I then had to track that and interestingly enough the data was such that what I was worried about was this thing was so bad that infact they almost stopped giving me data every day. If you weren’t so compulsive about this data we will not share our data with you because it your heartrate goes up. And then this is the one that really counts, when you have got an infection, this is called ESR roughly a sludge floating around in your bloodstream. So I had an infection and I was trying to take pictures and sure enough they weren’t changing but I convinced the doctor that they were changing, and he wasn’t looking at my pictures anyway. Until they didn’t get good enough I couldn’t get out the hospital, and after about two days in the hospital and hospital food you want out. One of the best investments I made in quantitative stuff is if you can do it is to go to Canyon Ranch. I me the medical director there and he looked at me and said you’re going to die and you’re going to have another heart attack, but I didn’t have one but I was going to have Type 2 diabetes so come and see me. I spent four days there and I got lots of interesting things like body scans and my girlfriend and now wife said, you can’t hear I said I know I can’t here. Look at how that drops off. You know that’s an age related problem and I can’t solve that but generally I can hear anyway in a telephone band and so what the hell. Anyway so I had to take that test and she said, okay, you may have sleep apnea. So I said, that’s a painful test to take, so I said yeah I’ll take a sleep apnea test; they wire you up and you try to sleep, so you have bad sleep for the night with all this apparatus on you so I had some of that. I think the key thing is wellness monitor which is basically why you want to monitor your ambient there. The one on terms of settling on stuff, I tend to now just use a pedometer; this nice little pedometer from Oregon Scientific. I weigh every day. I spend about half the time in Sydney and this is one of the walks I take. This is with a GPS, and then, but this is I think where we want to be which is essentially the cell phone is a health platform and want to know your GSR and stress as a function of where you were in life, you know, what meeting was in in or what’s going on here, and I think a device should have GSR. This is my favorite device. This is the old body media device. And this measured everything. It had and infact it had heart rat, but you can’t because I was wearing a strap at the time, and that device was this old one and it measured all that stuff. This is the device you want. They don’t sell it anymore. They don’t give you the stuff anyway and this is the one you get now. They have a good business model but lousy for us because all it has to do is with weight training. So all it’s doing is measuring your caloric output and steps. But this device and particularly with the heart if you can get heart rate on it and they can do that, is my ideal device.

Of course your weight depends on how much you eat. Here’s the problem for weight control; it’s not the output it’s the input and you have to measure all of this stuff. I did that for about a week to get an idea of what all the numbers are, and you have to be pretty anal compulsive to do this daily. But it would be great. If this thing could recognize stuff that would be the way to go.

About the presenter

Gordon Bell gave this talk.