Last night the first meeting of The Quantified Self Meetup Group took place at Kevin Kelly’s house in Pacifica. Here are my notes:
Kevin KellyÂ (KK) Magazine junkie science groupie host
Lisa Betts-LaCroix home schooling
Joe Betts-LaCroix now future earth people children
Nathan Myers software wind minerals
Lee Corbin software hobbies math chess history
David Kirby machine learning applications
Tim Ferriss athletic performance language acquisition cognition
Hor machine learning life extension
Graham Bullock product health effects
Dan Brown sleep monitoring for better rest
Paul Lindow film binaural language
Jerry Legert ontology unix physics
Peter Mortonsen design irrationality
David Duncan writer radio experimental man
Alex Bangs personal health
Alexandra Carmichael open-source health research
Daniel Reda open-source health research
Gary Wolf (GW) reporter utopian
Vivian Dissler new conceptions of health
Moor unwilling rebel leader
Rajiv Mehta tools for living
Seth Roberts sleep mood self-experimentation flaxseed
Ka-Ping Yee visualization democracy lambda graphics usability
Stephanie life-streaming ubiquitous ambient
Chris social constitution of self online
Steve complex systems accumulating evidence reflexively
Neal futurism python family jazz piano
Kevin Kelly introduction
Like to keep broadening it, before we get to the idea of saying we’re not interested in this. Using data & measurement for self-knowledge. Revolving around numbers and measurement. And measuring ourselves. Here we’re interested in science upon yourself.
Stuff people are interested in
Self-measurements of health
Using geocoding for self-analysis
How have you incorporated measurements into your life?
Where do you do your research? Where do you get your ideas?
When you have a lot of data, how do you extract something you can see from it?
What data people would be willing to share with others? What private?
Measuring the effects of measuring
Resources, other groups
Spheres beyond health – such as time use, learning, how quickly you acquire new items
How to measure learning
Aesthetics of self-documentation
What do to with the data that we gather
Can you learn too much about yourself?
Environmental data, brain scans
How quantified health can be used to collaborate w/ traditional medicine
The greater good of all this – what potential for doing good?
Stuff people are willing to talk about
I’ve tracked my time for 3 years – how to analyze
Things I’ve measured
What professional athletes do in this domain
Learning, how quickly you learn
Design principles for measuring oneself
Genes, brains, & body – the Experimental Man
Ka-Ping Yee: Time Tracking
Tracked use of his time. Felt like problem not making good use of time. By collecting data, would help. How much sleep. Where is my time disappearing? Wrote script that keeps little window open in screen. Types in whatever he’s doing. E.g., “at QS meeting”. Logged with time. [Question from audience: what have you learned?] not a lot. Hope to learn. Can measure how much sleep I’ve been getting. Turns out to be getting more sleep than I thought.
Roberts: this data useful if you change something. You can notice unexpected changes.
Ferriss: MeTimer: tracks web browser use
RescueTime: indicates active focus, & local applications. Fascinating. Making me more effective. Doing the right things.
Xobni: identifying email hotspots. Top 10 contacts.
???: made pie graph of where we’re spending our time & is it compatible with what we say we’re about? Do we spend our time on our kids, business, health?
Clickable menu to help track time.
Yahoo widget stopwatch. Tracking work hours.
Surprises: so much unstructured time, more time spent with kid bigger than expected. Agast to realize one project was taking so much time. I’m really doing a lot of things. Decided to focus my life more. Ultimately the most satisfying thing to me. Now I’m working on that company. Me figuring out where I’m putting my creative energy.
The challenge of assigning tags to things. Too much or too little detail. Just write in your own words.
David Kirby: time tracking mood, sent message at random times to cell phone, got a lot more honesty about activity. Extremely difficult to do in social environment. What are you doing, do you feel productive? What is the sampling period? And what is the associated honesty of the data? I became very afraid of wasting time when each message would come it.
Tracking leads to optimization. As soon as she gets people to track their diet, it massively improves. Helps with eating disorders. Nigerian lawmakers tracked by journalist. Just the fact of publishing it greatly improved voting records.
Tim Ferriss: Athletes
Recorded every resistance training workout since age 15. Now 31. Diet & so forth. Go back to notebooks to see how I looked in this or that photo. Related to performance enhancement. Athletes are v. interested in relative vs absolute measurements.Â Body fat percentage – many different methods. All have problems. Real benefit is in measuring progress – difference up or down. I measure VO2 max. BMI is waste of time. VO2max tends not to change much. Only tested twice. Very similar in spite of different regimens. Cyclists measure to see performance maximum – can I become professional? Lots of blood testing. Every 2 or 3 tests, plus saliva-basted testing. $500-$1000 per time. Complete metabolic panel. Cortisol, estradial. Done 1st thing in morning. [Someone else: 120 things I track.] Most athletes use a laundry list of drugs. They cycle off to compete. They look at ratios. E.g., free testosterone to epitestosterone. Steriods are like antibiotics. Many of them. Betablockers: used by Chinese archer, concert pianists, entire orchestras. Subject in expts related to body heat. When does your body shut down? How to best dissipate heat? Wanted the data, would have cost $100K. Found that my ability to dissipate heat is severey reduced. Surface areas of lungs compared to surface area of skin. Brain shut down at unusually low temperature. Heat stroke susceptible. Recovery time measure of fitness. Sets, Reps, rest periods. Within rest periods you can track several things. Return to basal body temperature, heart rate, muscle tonus (galvanic skin response). Don’t believe in general intelligence or general fitness. I measure sex hormones, metabolic panels, IGF-1, indirect measures of growth hormone, free testosterone, T3 T4 TSH insulin (wide diurnal variance). How much insulin is produced, how much remains after meal. Athletes only measure something if they can act on the data they gather.Â I don’t want 24/7 measurement – too much data. I take a lot of experimental supplements. I like watch the response of my autonomous nervous system. Ephedrine hydrochloride. Widely studied.Â Yohimbine HCL: proposed as safer alternative to ephedrine. Not a selective agonist. I wanted to measure my blood pressure very closely.
You can order blood tests online & go to local places that will draw your blood & send you the results. Don’t need doctor’s prescription. Look up symptoms for whatever would require that type of testing.
Any statistical tools that will help you calibrate measurements against each other?
Templates for analyzing small data sets: couldn’t find.
Statistics packages, such as R. Google Website Optimizer. Set up experiments for website visitors. Test different pictures on your page. Ramsey 2-color theory. The known Ramsey numbers. If you know there has to be a mutual group of 5 people that all know each other, you have to have at least 43 people in the room.
Suggests data tutorial for the self-experimenter. At future meeting.
If there were a easy to use tool set, if you wanted to run experiments, finding things that were harder to find or predictions. Good tool honeypot for data. Norbert Weiner talks about that.
RescueTime can take aggregate data of 30000 people, reducing data entry from 100 items to 2 items. Hope there will be communities that will follow protocol.
Gary Wolf: Learning/Knowing
Did long story for Wired about managing 5000-item dataset. Keeps track of what you’ve learned & haven’t learned. Makes prediction about what you need to practice something to remember forever. How deeply rooted it is. Street you’ve lived on when little kid. I’m measuring 2700 index cards. Spanish vocabulary. Modal sentences for grammar learning. Maintaining at 90% remembering. What 1000 words do you need to know? Super Memo. Pieter Wozniak, Polish computer scientist. When you do your tests every day, you get 85%-90% on every test. It shows you a set of cards. If you want to learn the most, maximum efficiency is 40% efficiency. Unpleasant.Â 5-point rating system. Implemented physically in the 1950s.Â Liter System. Pimsler implements spaced repetition in background way. Index cards a metaphor – all online. You want to learn constantly for short periods of time over long time spans. What’s the lifetime limit on what you can learn? Cabbies in London: larger hippocampus. There are some glitches: learning some things can interfere with knowing other things. Cards in Super Memo that often come up. You can figure out why you’re not learning them.Â Everyone knows that poor sleep impares learning. How much will your performance decrease? You need a lot of sleep data & lot of learning data. We’re on edge of being able to do that.
David Duncan: One of brain scan tests for my book has to do with memory. As we age, we don’t forget things, we have inability to filter out. Series of faces & landscapes. Then forget faces or landscapes. Can you filter out what you’re not supposed to remember? Compare young and old. Older people can remember as well as young people but can’t filter as well. I came out right in the middle.
I’ve been doing tai-chi for a number of years. Doing for 5 min/day better than 1 weekend/month.
How do you measure learning? Not just via standardized tests. Cognitive & batteries of tests that people take. Tell you what your brain age is. Column for Portfolio. Cognitive drug research. Did better than I expected, didn’t want to take again. On website. On wine, my timing was way off. Check comment stream, people who got younger thought it was a great test. Issues with speed of server.
Rajiv Mehta: Ways This Helps
Helping people take care of their health. Individual perspective. You have some regimen in mind: chicken soup or whatever. What gets in the way of people ttaking care of their health? We forget a lot, life gets busy. Keeping detailed records impossible for vast majority. Made prototype device. Gives reminders for thigns that are scheduled. Makes easy to record symptoms moods, etc. anything you think is important health. User tracking exercise. ZumeLife.com. Tracking her carbs, glucose, medications, symptoms, shortness of breath. Speak into it. Transcription service. Things that don’t lend themselves to charts get put into notebooks. USB connected to your PC. Every day or so. NY Time article about 42-year-old person w/ diabetes. Complicated case. Lots of drugs. [can you use any category?] Right now. 5 categories: meds, biometrics, exercise (name, duration), food (carbs, points), health-status questions (how much coughing, how did you feel). Most been using: 7 months. One person had sclera derma. Intense pain, cold hands. She started measuring her pain levels, realized she was having 3 or 4 intense episodes per week. With her printout, she & her doctor said: you have a problem. Over 6 week period she went from 3 or 4 intense to Â Â Â 1 mild episode/week. About $40-50/month. Can’t ask people to buy a special phone. I’m not waiting for United Health Care to say it’s interesting. Institutions don’t want to deal with smart phones, they want a special device. We’ve found everyone experiments. If anybody is going to figure out what works for her, it’s her. Doctors benefit from charts. The experimenter is her. Our doctors are at best good advisors. We cannot abdicate responsibility.
Peter Mortensen: Motivating Running
If you want to help people who are not enthusiasts. Need proud & constant presence. Nike Plus Sports Band. Wear all the time I’m awake. Not just a string around my finger. I can see how much running I’ve doine in last 3 months. Also have my last run. More successful than Nike Ipod. You should hide the change you want to make. Also USB key, syncs/ with NikePlus website. Graphs, etc. Even cooler it charges battery when plugged in. I only hear I’m interested I’m wearing a nice watch, never that I’m self-tracking. Chip on shoe, Bluetooth. Impact accelerometer. It’s an attractive watch, I’m proud to wear it. Aesthetic thing. An ugly watch would not have the same effect.
Raj: If you’re sick it means something bad. If there’s going to be a device they have to be proud of it. Before Ms. L got our device she was trying to use her Treo. It got in the way. It’s rude to answer your phone if you’re having a conversation. People would get upset. This device: It’s Ms. L taking care of herself.
Nice scales for tracking weight. Cable going off somewhere. Body composition Bluetooth scale. Tonita.
Seth Roberts: Sleep
Early self-experimentation about acne. Found out stuff doctor didn’t know: really useful. Maybe would help with sleep problem: Waking up too early in the morning. Tried many things for 10 years, everything failed. All ideas were wrong. Finally got a new idea due to analyzing data, noticing less sleep when lost weight, told class, student told him about another way to eat less, required changing breakfast. Change in breakfast caused early awakening to get worse. This was the first thing that had made a difference. Led to discovery that any breakfast hurts; supported by rat research. Later found that standing a lot improved sleep. Had to stand at least 8 hr to get effect; no effect of 6 hr of standing. Great sleep when stood 10 hr but really hard to do. More recently discovered that standing on one leg to exhaustion helps. Do twice in one day. Same effect, roughly, as many hours of standing but only takes a few minutes. Now working out the dose/response function.
KK: looking for people who want to blog on the blog itself. If you’d rather just share a comment or idea just mail it to us (GW & KK).
GW: lot of dry timber waiting for a spark here. Value of this meeting, can something bigger grow out of it?
Bodymedia.com making their gadget available to retail. Measures body temperature, galvanic skin response. $399 plus $20/month. 2. Body computing comfernece at UCSF Oct 25. internal 3. 23&me has dropped their price. Version 2. 1/3rd price.
Send stuff to KK & GW to post on blog.
More Gary Wolf’s account.