3D model and animation by Jeremy White via NYT.
What I learned from my Uber data
I save $703/month using Uber versus owning 2 cars (🚙🏎). I’ve spent $12,041 on 686 Ubers over 6 years. By comparison, car ownership cost $11,783 per year, most of it depreciation. What could $703/month buy you? Well, four years of $703/month and you’ll save enough to climb Mount Everest 🏔 unguided. Or you could buy a shiny new laptop every 3 months. More details below.
My uberPOOLs average 1.6x longer ⏱ and are 27% cheaper 💵 than UberX, and this bargain is worth it for me much of the time. Because of UberPOOL, my average Uber in 2016 was just $11 per ride. By comparison, I spent $37 per ride on average in 2011 and 2012. It’s as if UberPOOL is paying me $28.50/hour to sit in pools, versus taking UberX. I spent 3.2% of my non-working, non-sleeping hours in 2016 in Ubers, almost 4 days! More details below.
Uber pays me $28.50/hour to sit in uberPOOLs, versus taking UberX, the equivalent of two simultaneous minimum wage jobs. More details below.
Only 4% of my drivers are women 👩🏻. I got a female driver only once every 25 rides.
Photo credit National Geographic Magazine, November 2016.
Earlier this week I blogged about 5 simple questions you can ask to determine AI hype.
Today I saw something with more hype than I would expect from a respected tech company. “eBay determines this price through a machine learned model of the product’s sales prices within the last 90 days.”
In my opinion this price prediction is not machine learning. It’s just math. It’s not a machine learned anything. It fails 1-5 of Stephen’s tests.
As Machine Learning captures our imagination, it’s important to separate the material from the hype. Here are 5 simple questions you should ask to help reduce AI hype:
- “How much training data is required?”
- “Can this work unsupervised (= without labelling the examples)?”
- “Can the system predict out of vocabulary names?” (i.e. Imagine if I said “My friend Rudinyard was mean to me” – many AI systems would never be able to answer “Who was mean to me?” as Rudinyard is out of its vocabulary)
- “How much does the accuracy fall as the input story gets longer?”
- “How stable is the model’s performance over time?”
source: Stephen Merity of Salesforce
My commentary on Stephen’s article. Stephen is placing the bar for ML high. AFAIK, Google DeepMind’s AlphaGo fulfills these five requirements, but Google Photos does not, since Google photo uses human crowdsourced labeling of what a cat is to determine cat videos. The fifth requirement, “How stable is the model’s performance over time?”, seems to be the least interesting and the most interesting at the same time. It’s the least interesting requirement because the evolution of a cat will indeed eventually change the way the cat looks and untrain even the most sophisticated ML model, unless the model adapts as the cat evolves. It’s the most interesting requirement because I can imagine how a model would fail to understand meme’s such as “Bye Felicia”, “On Fleek”, “idk my BFF Jill” or another human-created that seems like gibberish to machines or people not in the know.
Skip to 3:01 to learn how the automobile differential allows a vehicle to turn a corner while keeping the wheels from skidding. It’s brilliant product marketing, using language and concepts to conceptualize the product features into something that everyone understands and wants to buy.
Here’s the recorded video of a 5-minute talk I gave at Battery San Francisco about Fix Maps. The talk format is called “Ignite” which calls for 20 slides that auto-advance every 15 seconds.
Here’s the slides from the talk.
Snow caves are fun to build and provide warm places to sleep and take shelter from a storm. This is a brief how-to guide to building a snow cave.
The snow cave in this example was built by 4 people near Skinner Hut at the edge of the timberline in late-December 2015 at 11,620 feet. Builders were Brett Poulin, Chris, Nick, and me, Neal Mueller. The cave we built was large enough to sleep and provide eating quarters and shelter for 4 people. It included a vapor escape for cooking.
This guy analyzed 250 SaaS pricing pages — here’s what he found:
- The average number of packages is three and a half
- 50% highlight a package as the best option
- 69% of companies sell the benefits
- 81 percent organize prices low to high
- 38 percent list their most expensive package as ‘Contact us’
- The most common call to action is ‘Buy Now’
- 36 percent don’t use a contrasting CTA color
- 63 percent offer a free trial
- 4% of companies offer pricing on a sliding scale
- 81 percent of packages are named
- 6% show a money back guarantee on-page
Read the full report.
Best definition I’ve seen of growing products on the web.
A growth hacker is someone who has thrown out the traditional marketing playbook and replaced it with only what is testable, trackable, scalable…while their marketing brethren track vague notions like branding and mindshare, growth hackers relentlessly pursue users and growth (source: Ryan Holiday).