This is a worthwhile short history of AI and deep learning, focused around the humans (Minsky, Hinton, et al) who brought us its top advances. Probably not earth-shattering for folks deep in the field, but entertaining and sets up a decent overview of the personalities involved. Skip the sections on AI ethics; they already haven’t aged well and I predict they will continue to age poorly.
I found this New Yorker longread super-troubling, especially with this setup early on:
Anyone working with data on a day-to-day basis should be worried about selection bias, randomness, or both when looking at such a diagram.
The overall article is highly recommended. I’m not sure what to think about it after finally finishing it, other than the feeling like humans continue to be terrible pattern-matchers, searching for meaning where there may be none. I tried Googling for other sources but this was the least salacious source I could find, perhaps because of British contempt-of-court laws. I wonder what else has been written that’s worth reading?
Absolutely amazing piece in the New York Times Magazine today by John Jeremiah Sullivan, about tracking down two female blues singers from the 1930s. Stop wasting your time on whatever it is you’re currently skimming and read it right now.
I also very highly recommend Sullivan’s book, Pulphead—a great read from start to finish.
“What happens when you hear a text rather than read it? The obvious thing is that you can do something else with your eyes.”
- T.M. Luhrmann
I’ve always been curious about audiobooks but my primary activities—working (i.e. coding), reading, making music—require my full attention. When I’m really working, even regular conversation can be a huge distraction. And my longest commute is a 10-minute walk, which offers little opportunity for listening.
I wish I had a hobby that required my hands but not my conscious attention. I know I’m missing out on a lot—not just audiobooks but NPR and other podcasts that I hear about from friends. Theoretically, I could sit and listen attentively to NPR or a book, but somehow I never make the time.
Maybe I’ll try listening to This American Life next time I’m cooking a big meal.
My first startup job (at HOTorNOT) was essentially a “startup school.” Though the site seemed frivolous on the surface, the team behind it was small, dedicated and fiercely data-driven–using homegrown tools—before being data-driven became a “lean” mantra.
These days, anyone can collect data simply by dropping in tracking code from Mixpanel, Heap, or even Segment.io, a meta-analytics service. But all this data collection is pointless unless you can use it to make decisions.
Collect correctly
The first requirement for data-driven decision-making is good data. Collect it all and make sure it’s correct, that you’re measuring what you intend to measure. It’s easy to make mistakes here—whether it’s a database query that’s slightly off, or missing a segment in Google Analytics.
If you and your colleagues don’t have faith in your data, then you won’t be able to make decisions with it.
Ask good questions
After you’ve got good data, you should ask good questions. The more specific your question, the better—not “how many users have signed up”, but “Which channels have high-value users come from in the last 6 months?”
This question is slightly more complex than it sounds—you have to define what “high-value” means (and different people in your company might have different opinions)—but it’s certainly answerable, and will give you more actionable analysis than just asking about “all users.”
Analyze and Act
Next, analyze your data to answer your question. There are plenty of tools available for this—segmentation, funnel analysis, cohort analysis, plain old number-crunching in Excel. I won’t go into detail here. But you should be able to answer the question you’ve asked if you’ve collected correct data.
The last step can be the hardest—acting on your analysis. If you’re like most product designers, you’ll often find data that disprove strongly-held opinions about your product. You may have to convince your team that your analysis is correct.
There will always be situations when you don’t have enough time to rigorously collect and analyze data, or when you are creating something entirely new—these are times to use your intuition and make predictions based on previous experiments. But if you have the time, making an effort to drive your decision-making based on data will pay off.
New York City lacks the startup density of Silicon Valley, but there are plenty of reasons to start your next company here:
A great (and growing) startup scene. Soho, Union Square and the Flatiron district form the new Silicon Alley, where you’ll have your pick of companies, meetups and events. Great startups are doing great work here across industries—including tumblr, foursquare, 10gen, Betaworks, turntable.fm, Etsy,Skillshare, and Harvest, among others 1. New York also has a sizable angel and VC community 2.
Bubble-free. SF often feels like a company town; in contrast, New York is a national hub for publishing, advertising, fashion, food, and finance, to name a few. Being in the center of all these industries is inspiring, and your startup can take advantage of this diversity.
The singles scene. Some folks might rank this as #1; I’m trying not to assume too much. But if you’re a single straight guy or gal in SF or Mountain View, you (or a friend) have likely complained about the singles scene in the Bay Area.
Forbes ranks New York as the best city for singles, putting it above San Francisco, which clocks in at #7 3. For single guys, the numbers don’t lie—there are 210,820 more single women than men in the New York metropolitan area 4. And even if you’re happily married, being in a city full of young people provides great energy. Also, being a tech entrepreneur feels more unique in NYC.
Official support. Mayor Bloomberg wants to turn New York into a tech hub, and city officials reach out to local entrepreneurs. The NYC Economic Development Corporation offers great resources including info on city incentives and discounted office space for startups.
The city never sleeps. NYC is much friendlier to the late-night hacker’s schedule—subways and buses run 24/7 and great restaurants stay open till 2am or 4am. Contrast this with San Francisco, where it’s hard to get good food after 11pm, even on a Saturday night.
The upstart startup hub. This one is hard to quantify, but there’s something exciting about being part of the NY startup culture because it is less established than SF—you’ll find plenty of room (and enthusiasm) for starting a new meetup, for example. There’s an energy in the air here, a sense that we are all part of something new and growing and awesome.
If you’re living in SF or elsewhere and looking for a change, come visit NYC and see if it’s right for you.
I’ll even buy you a coffee.
See a bigger list of NY-based companies here: Internet Made in New York City.
The New York market on AngelList hosts an excellent list of NYC-based investors and angels.
Source: Forbes
Source: Boston Globe
Here’s another easy win—never make users log in to unsubscribe from email notifications. For nearly every website, there’s no security risk in allowing users to unsubscribe without logging in, and there are many benefits, including goodwill from users and fewer spam reports. Unfortunately, most popular sites don’t take the simplest approach to solving this problem.
Bad approach: Twitter & Facebook
Twitter sends a lot of different kinds of notifications, and their approach to notification management is complex—as evidenced by their incredibly complex footer text, which wouldn’t fit into 2 tweets, let alone one.
Twitter requires you to log in to change any of their notification settings, and even after logging in, they don’t give you a simple way to unsubscribe from all email. You have to trawl through several dense pages of settings.
Facebook’s approach is similar—they require login, then you have to figure out which of their many notification types to disable.
Good approach: Yelp
Yelp does a great job. They let you manage all email notifications without requiring login—just click through from the email and you get a clean management page that works with just your email address.
Good approach: MailChimp (and Feedburner)
MailChimp and Feedburner (used by many newsletter senders) provide instant unsubscribe with a simple option to resubscribe if you change your mind. Clean and respectful.
Doing it the right way is easy
It’s easy to choose a UI pattern for your unsubscribe links.
If your service sends many different types of email, follow Yelp’s lead — let users click through and manage all notification types without logging in. If you’re sending a newsletter, let them unsubscribe with one click.
Make the right choice. You’ll make your users happier and reduce your spam reports.
We’ve been making interactive web applications for since at least 1995, when Netscape shipped JavaScript (née LiveScript). Yet I still see the same basic usability mistakes on many websites.
I’ve started a list of the ones I think we should all agree on—a set of UI Rules, if you will. First on the list is one of the easiest to fix.
Rule #1. Focus on the first text field.
Most important for login and signup pages, but this rule applies to almost any form on any page (with the possible exception of pages with lots of separate forms). Why not save your users a click when they’re performing the primary action on a page?
Let’s look at some examples. I don’t mean to pick on any of these companies—my methodology here was picking randomly off the list of trending startups on AngelList.
The Fix
Developers often use JavaScript to solve this problem, but there’s an even easier solution—the autofocus attribute on input fields. It’s supported in all modern desktop browsers, but not on mobile (potentially for usability reasons according to this explanation from Wufoo).
Let’s all agree to follow this rule from now on—however you do it, focus on the first text field on the page. In almost every case, you’re saving your users time, which will lead to increased engagement and conversion rates. Together, we can make the Web a less frustrating place.
It is much easier to start a web startup in 2011 than it was in 1995, or even 2005.
You don’t need upfront capital to buy racks of machines, miles of CAT-5, routers and load balancers, redundant power and backbone connectivity — all are available on-demand from Amazon for pennies an hour.
You don’t need to write server-side Web frameworks — you have your pick in nearly any language. In fact, you have your pick for any number of services you might once have had to develop or maintain yourself: email delivery, web analytics, A/B testing, server monitoring, file hosting.
Because the infrastructure cost of starting a startup has gone down, the amount of money you need from investors to get your company off the ground has also decreased. You can build a prototype in a few months (as every YC class demonstrates, twice a year) for less than $20,000.
And there are more people to give you that money — VCs, ex-Googlers and Facebookers, other angels, startup accelerators and incubators.
But ask any startup around what they really do need right now, and the answer is clear — great technical talent. Startups of all sizes are competing for a small pool of highly skilled candidates, and in startup hubs like San Francisco and New York the competition for engineers is fierce. This scarcity has persisted despite consistently high national unemployment rates.
And because it has become so easy and cheap to start a startup, convincing great engineers to join an existing startup has gotten much harder.
All of these trends will continue — infrastructure costs will continue to decrease, tools for creating web products will get more sophisticated, and more people (and more kinds of people) will be investing in startups at all stages.
But those startups who can hire and retain great talent will achieve greater success because they have the capacity to out-execute their peers.
…when you are young, you think you can predict the likely pains and bleaknesses that age might bring. You imagine yourself being lonely, divorced, widowed; children growing away from you, friends dying. You imagine the loss of status, the loss of desire—and desirability… What you fail to do is look ahead, and then imagine yourself looking back from that future point. Learning the new emotions that time brings. Discovering, for example, that as the witnesses to your life diminish, there is less corroboration, and therefore less certainty, as to what you are or have been. Even if you have assiduously kept records—in words, sound, pictures—you may find that you have attended to the wrong kind of record-keeping.
Sense of an Ending is a short and wonderful book, a slim 180 pages readable in a single sitting. There’s much to savor, though, and like Marilynne Robinson or Alice Munro, Barnes chooses every word carefully and precisely. Highly recommended. Amazon