The Gong Show

Here’s a question I’ve been tossing around for the past week: Which web service has the best representation of the In-Real-Life (IRL) social graph?  Knowing who is *actually* friends with whom IRL (and not just accepting a friend request out of social guilt or apathy) would be a valuable data set to own. The IRL social graph has existed for as long as humans have formed friendships, yet throughout that history I can’t think of any really high quality historical attempts to map it. I’d go so far to say the IRL social graph is so important that the social graphs of many popular web services are valuable based largely on their ability to proxy the IRL social graph.  So, who has the best proxy, or perhaps even a totally accurate model?

Lets get the obvious statement out of the way: it’s not Facebook. If I look just one degree out on my Facebook social graph I see people I don’t even recognize anymore, and I think that’s true too for many of my friends using the service. Some people (not me) will accept a Facebook friend request after meeting a friend of a friend out at a party.  Perhaps this is good social graces in our modern times, but it’s a big part of the reason why Facebook’s social graph is not a good proxy to the IRL social graph.  The IRL social graph is likely a subset of FB’s graph, which means the edges in the IRL social graph are more meaningful than in FB’s graph.

When Facebook Places launched, Nate Westheimer mentioned in a well-thought out post that Foursquare owned the IRL social graph. Foursquare’s sensitive data asset (the locations of my friends and me) and their respect for users’ privacy encourages me to permission information to only people in my IRL social graph.  I believe Nate and I agree on that point. But, Foursquare isn’t popular enough to own the IRL social graph yet because most of my IRL friends are not on Foursquare and have no interest in Foursquare (yet). So, Foursquare is probably well-positioned to own the IRL Social Graph in the near future, but they’ve still got a long road ahead of them.

Along the same lines as Foursquare, I think Blackberry Messenger (BBM) has a solid shot at capturing the IRL social graph, if they weren’t limited to Blackberry-only.  When I carried a Blackberry, only my IRL friends got BBM access from me, and so for a brief window when many of my friends carried Blackberrys, they had an interesting data asset.  But now almost no one who is important in my life carries a Blackberry (at least not as their primary phone), so it’s no longer a good social graph to own, and not a good map of the IRL social graph.

How about my cell phone and text message logs (aka Google Voice)?  If I call someone, they are very likely to be in my IRL social graph because I reserve phone calls for only people that merit such a high level of attention.  If I were in high school and were a user of Google Voice, I’d say this is one of the best proxies to my IRL social graph, but Google ends up with noisy data because I use my cell phone for work all the time, which clouds my IRL social graph.  Google Voice doesn’t know if a person I call often is an old college roommate I’m reconnecting with or a newly minted MBA that I’m phone-screening, and the difference between those two types of calls is essential to mapping the IRL social graph.

A good picture of the IRL social graph would understand the difference between my different sets of connections… it would know that work friends, entrepreneurs, long-time college friends, family, and others are clumped into groups that rarely intermingle. In fact I often actively try to keep these groups separate in my life. Most social services, especially Facebook, really fail at capturing this aspect of the IRL social graph. I know Paul Adams did a great job explaining this concept in his presentation on real world social networks, and I completely agree with him on this point.

In case you can’t tell yet, this post is an open question (please fight for your winner in the comments). I haven’t figured out who has the best proxy for my IRL social graph.  The company with the best chance of figuring it out, for me at least, is Google.  Google has my search history, all my email, my calendar, Google Chat logs, the docs I’ve shared with others, and (if I weren’t an iPhone user) my phone logs and txt messages. From that mess of data, I could easily see a great map of my IRL social graph emerging. But, Google doesn’t *get* social (need proof? See Orkut, Buzz privacy issues, Wave, Lively, Dodgeball etc…). They would have to fight their own instincts the whole way towards mapping the IRL social graph correctly.

Applications on the web can be classified into archetypes.  Once you look past the UI and dig down to the organizing principles, the there are more similarities than differences between services as disparate as TwitPic and Mahalo.

CRUD Apps: CRUD stands for Create, Read, Update and Delete.  This app archetype covers the set of applications where the UI is really just a layer on top of a database designed to store inputs in a predefined structure, and then recall or manipulate those inputs as desired. Many apps fall into this model, such as: del.icio.us, flickr, gmail, wordpress, . If I had to defines this class of apps in a single layman phrase I’d call them “published storage.”

Transactional Apps: This category covers marketplaces, payments, ecommerce, online classifieds, event, etc… and any other type of app that enables the transfer of goods or services between parties. Some of them are more aggressive than others about getting in the middle of the transaction (with something like StubHub on one end of the spectrum and craigslist on the opposite end). The commonality is they all facilitate transactions.

Aggregation Apps: This category of apps is a second-order layer of applications that often ride on top of a set of first-order CRUD apps. This is an app that collects information from first-order apps by some combination of crawling/scraping, user-generated submission, and business development.

Aggregation apps only make sense when the underlying sites being aggregated are highly fragmented and do not share common data.  For example, aggregation works great in job boards: Indeed and SimplyHired aggregate many hundred niche job boards around the world successfully because the underlying job boards are highly fragmented and do not share many common jobs.  Aggregation does not work in online auctions.  That’s because the market for online auctions is not highly fragmented. eBay owns 90%+ of the online auctions, so if eBay opt-ed out of being aggregated then your business is toast. 

Read-Only Apps: There are a set of apps out there that look more like TV than anything else.  They are read-only apps and take no input from users.  Much of the original brochure-style websites in the 90s followed this model.  ABC.com is a great example of what I’m trying to describe in this category, and if it was not possible to rate shows or create a personal queue, then Hulu would be a good example of this type of site too.   Because these sites are read-only, one could argue they’re not applications at all, but some of the biggest sites on the web follow this model, so I wanted to include it in this list.

The list of archetypes is currently neither mutual exclusive nor exhaustive, but I wish it were. If you think this can be improved upon be either reclassifying or adding more categories, then don’t hesitate to pipe up in the comments.

I attended the Betaspring Demo Day down in Providence, RI yesterday.  Congrats to all the teams who presented.  I thought the overall quality of all the presentations was really high, and I’m delighted to startup culture blossoming so nicely in Rhode Island. Scott Kirsner has a great summary of the day and all the demoing companies.

Two Comments (only tangentially related to the Betaspring event):

1) I must have been not paying attention to local events/media because I only learned about Betaspring for the first time two days ago (hat tip to Jon Pierce, thanks!). That’s my fault; I should have been listening more closely to what’s going on in the area this summer. But, this lapse in attention made me dig into just how many accelerator programs there are now, and this semi-complete list of 100+ programs is staggering.

What’s even crazier is each of these programs is producing around 10 new companies a year, if not more. That’s around 1000 new companies a year. There should be a whole wiki dedicated just to keeping track of them all. Crunchbase sort of does this, but they don’t have a differentiation between “investor” and “accelerator” in their taxonomy. And, what I’d really like is a schedule of upcoming demo days in these accelerator programs, which Crunchbase certainly doesn’t provide.  I would love it if someone built a resource dedicated to tracking all these programs and their progeny (and nothing else). I’m sure some ambitious VC summer intern has done this work already at some firm, but I’d love to see this data get crowdsourced.

2) I love show-and-tell. I’m a sucker for any pitch that involves less talk and more demo.  I love the way people get emotional (excited, bashful, optimistic, etc…) about what they built.  I love how raw alpha products can be, and I love the feeling of seeing a really compelling product peeking it’s nose out from behind the mess of awkward, early UI and bugs.  That’s the best part about attending these accelerator program demo days: they are pure demos, back-to-back-to-back.

I went over to the Seaport District yesterday to visit the MassChallenge office space.  It was an impressive site. Lots of entrepreneurs running wild, networking with whomever walked off the elevators, all collaborating on each others projects.  The energy was infectious.  If you’re a VC in Boston focusing on new media, I highly recommend taking a run through the space over there.  I know I’ll be back for more.

I was surprised to learn they are a non-profit.  Along the same lines, they take no equity in any of the startups involved in the program.  For an aspiring entrepreneur, that’s best deal I’ve heard of out of an accelerator program, considering both the physical resources and mentoring support the program provides.

MassChallenge took over the 14th floor of the One Marina Park building down next to the ICA… it’s a beautiful space. You have 360 degree views of the waterfront and the city.  And watching planes take off and land across the water at Logan is fun. It’s a far cry from the typical hole-in-the-wall spaces 95% of startups inhabit.

Thanks to Akhil Nigam and Ty Danco for showing me around. I wish all the startups there the best of luck.

I attended the Capitalize event thrown by DartBoston last night. It featured Google Ventures (specifically, Rich Miner) assessing the pitch by Hello Vino.

The most interesting point of the night came when Rich described his favorite way to measure Active Users: the 30-30 statistic. Most startups measure their Active Users based on some activity over the past X days (where typically, X = 30). By contrast, 30-30 is where Active Users are measured by counting all users who have used the product in the last 30 days who also signed up longer than 30 days ago.

The purpose behind the 30-30 methodology is that everyone who signed up in the past 30 days is an Active User by definition in the more common 30-day window definition of Active User, which is needless noise when you’re trying to determine engagement. 30-30 is a more pure measure of true engagement that doesn’t include people that visited your site once and never returned… it’s a great way to cut out the noise in a typical 30-day Active User window. The people who are left in 30-30 are the ones that are truly engaged.

Side note: the event was held in the SCVNGR offices, which are beautiful! Nicest startup offices I’ve seen in awhile (of course, I’m used to seeing NYC hole-in-the-wall spaces for startups, so I’m not the most unbiased viewpoint on this subject).

I’ve been playing around with Manu Kumar’s CardMunch product, and it’s really inspiring.  Simple proposition: take a picture of a business card with your iPhone and get a perfect scraping of the information imported into your contacts.  How can it be perfect? It leverages Amazon Mechanical Turk (MTurk) to have humans do the translation cheaply enough.

This creative use of MTurk to solve an otherwise very difficult OCR problem is a great example of the possibilities Human Computing offers, and I think not enough entrepreneurs are thinking of MTurk (or it’s competitors like CrowdFlower and Samasource) as the answer to the difficult engineering challenges they face.  

There’s a set of problem that can be labeled Turkable Problems. The problems that engineers have used MTurk to solve today are certainly in that set. But, also in that set are problems we have not yet thought to use MTurk to solve. Turkable Problems is a far bigger problem set than we suspect, and innovation will move very quickly as more engineers turn to MTurk in the future.

I’m concerned I’m losing touch with the behaviors of a typical internet user.  It’s arguable whether or not a “typical” internet user even exists… but when I use that phrase my intentions are to get at:

• What needs does the internet fill for people, on average?
• Where do they go to fulfill those needs?
• How long do people spend online, on average?
• What are the means of access?

I know these questions can be answered quantitatively with service like comScore or industry reports from people like Mary Meeker at Morgan Stanley, but data alone is really insufficient to really feel in touch with how people use the web. This is my concern… I know the data and read it regularly, but I have not spent enough time with users directly lately.

In college I took a few classes that involved performing contextual inquiries into specific professions and populations to do needs finding and generally explore usage without having a predefined thesis.  I learned more from these exercises than I did from entire classes in other departments. Seeing people actually try to get something productive done via a web service can be a painful process to watch… You have to grab your own wrist to hold back from taking over control and helping people find what they’re looking for more simply.

Similarly, at the Web 2.0 Summit a few years back (I think 2007), Safa Rashtchy held a session where he brought in five “average” internet users on stage for an interview with the audience.  The results were really eye-opening… stuff like using ask.com for search or not knowing you could get driving directions online.  It really helps provide a solid frame of reference to do contextual inquiries and user interviews like this regularly.

I’d love to see a regularly held event (maybe once every other month) where a professional user testing analyst pulled in 8-10 users off the street (or likely Craigslist) and asked them about what they do online.  There would be no purpose (not user testing a specific site or feature).  It would be strictly exploratory. I don’t think it would cost much to put this event on, and I think the outcome would be both enjoyable and surprising.

Oracle has sued Google for infringing both copyrights and patents associated with Java, the development language and platform Oracle acquired when it bought Sun. Google built Android on top of Java, and now Oracle is suing Google for the way it is using Java-related intellectual property.

The obvious headline “Oracle Sues Google” is a red herring. The outcome of this lawsuit is irrelevant at this point. The damage is done, and Oracle already lost. The real story is “Oracle Spooks Their Development Community.”

When Oracle acquired Sun for $7.4B, the two most important assets they got were MySQL and Java. These technologies are platforms that millions of developers and companies around the world depend upon. This ecosystem of developers is what makes these platforms so powerful… this network of developers help each other tweak their MySQL configurations, build drivers to interface with new languages, create modules that solve common problems, and build open source Java libraries that benefit each other. 

Now that Oracle has shown that they’re ready and willing to sue developers that leverage Oracle-owned technologies, developers will lose faith in these platforms. Today, many intro to programming classes are taught in Java because it’s generally approachable and intuitive to non-programmers.  I’ll go on record and make a bet that 5 years from now, the % of intro programming classes that teach Java will be significantly less than today because the teachers of the world will view Java is a decaying platform. Oracle is squandering their ($7.4B expensive) asset, and by suing Google with the Java patent portfolio, Oracle is accelerating the pace at which they will lose this rich ecosystem of altruistic developers.

I know that Oracle thinks they’re doing the right thing. The motivation for the lawsuit is that Google is forking much of the work that Sun developers did in building J2ME by building Android UI libraries on Java that replace J2ME functionality. Building Android on top of Java both undermines Sun-developed Java libraries while also standing on the shoulders of Java’s network, attracting developers that are already familiar with the language.

But Oracle is wrong. They should be thanking Google for keeping the Java language alive and relevant in a world where programming languages and platforms are a dime a dozen.  By building Android on top of Java, Google is helping Oracle extend the Java platform to a wider audience.  Oracle sure has a funny way of saying “thanks.” The lawsuit might generate near-term wins like licensing fees or injunctions on Android development, but it represents a long term failure of vision by Oracle.

After four years in VC, I’ve met a bunch of investors that are optimists and a bunch that are pessimists by nature.  At the risk of over-generalizing here, I’d say that optimism is an essential quality for a VC. Every deal will always have some hair on it on, and you’ll always find more hair the harder you look.  To make an investment you have to be able to look at the deal as a whole, embrace the hair, imagine “what if…”, and then fall in love (again and again).  That process is just too difficult for someone who is a pessimist at heart to do repeatedly. 

Charlie O’Donnell wrote a post yesterday entitled Changing The Wrong Ratio. He makes many points across the post that tie together well, but I want to pull out one that has been banging around my head for the past week: When there are already so many great startups going on, why are so many hackers strongly inclined to start their own company instead of joining a startup that already has some momentum and/or funding? David Tisch deserves big credit for setting my hamster wheel in motion on this topic last week.

To start, let’s get an important assumption out of the way: Now, more than ever, hackers are inclined to start their own company instead of working for someone else’s startup. If you don’t believe that’s true, then voice your dissent in the comments or forever hold your peace because for the rest of this post I’m going to assume you agree with me here. My evidence is strictly anecdotal (aka non-existent), but it feels very true, particularly in line with the explosion of seed cap funds/funding.

Charlie’s absolutely right that a startup’s biggest challenge is recruiting, so that’s why this question of Start VS Join is so important. As more entrepreneurial effort/HR fragments across a wider number of startups, a vacuum of talent forms, which sucks the oxygen out of the room for all startups.

So, why would a hacker rather work on their own startup than join a hot, growing startup? Some hypotheses:

• Equity alignment: a 1% piece of a hot, rapidly growing startup is a generous grant for a great hacker, but feels flimsy when compared to owning 60% or more of your own idea (after bringing on some other founders). You only have to be 1/60th as valuable as super-hot-startup to talk yourself into this logic, which isn’t a great argument, but definitely plays into people’s thought processes.

• Be your own boss: it’s not easy to work for someone else. For it to work out at all, you have to really respect your boss, and even then, having someone you report to at the end of the day can rarely ever compare to being your own boss. Following your passion everyday instead of following others’ orders is very alluring.

• Making your dent in the world: someone else’s startup is likely not going to give you the same satisfaction of creating an impact when compared to working on your own idea. With the exception of arbitrage businesses, startup have visions of making a dent in the world beyond just making money. Unless you’re 100% behind super-hot-startup’s vision, you’ll feel more motivated to shape the world using your own idea more than someone else’s. It’s so much more satisfying to scratch your own itch: especially when the need your company solves is also originally your own need.

• The allure of being in charge: many people aspire to run the big show. I think this common urge is part of the reason TV shows like The Apprentice, Undercover Boss and Shark Tank have success — don’t forget Who’s the Boss and Charles in Charge ;). But seriously, we live in a culture that, rightfully, values leaders, and so starting your own company is a fast way to rise up to the top, even if in the beginning you’re the leader of a company of one.

• De-risked (historically speaking): starting a company is crazy risky and very likely to fail. But, failure today is cheaper than ever because startups are so capital efficient in the very early stages. The causes of capital efficiency in early stage web services have been outlined many times before, so I won’t rehash them here. The key is you can now fail and iterate more quickly and cheaply than ever before, so the negative reasons against starting your own company have been reduced further than before.

I’ve met many entrepreneurs that are in love with the idea of running a company, and I think a number of the motivations I outlined above are part of their thoughts. I really appreciate their passion, but coincidentally, these people don’t often overlap with the set of entrepreneurs that impress me the most. The entrepreneurs that impress me the most are the ones that are completely rabid about solving a need, and they happen to be starting a business because it’s the most direct way to do so. When the going gets tough in a startup (as it always does), the people who love the idea of starting a business don’t love it as much, but the people who have an insatiable urge to solve a need are unflappable in the face of adversity.

I’m sure I’m missing other motivations to start a company, so please correct me. I don’t think having a plethora of startups is necessarily a bad thing overall, and I’ve argued before that there’s no such thing as too much seed capital for new ideas. But, I am concerned about a talent shortage for existing startups, and I hope that teams who share very similar visions can find ways to work together, perhaps even merge as they grow, so in the end we can build the lasting digital skyscrapers of tomorrow.

(PS: I can’t believe I just pounded this out on an iPhone keyboard. Next time I blog away from a comp, I’m just going to reblog that crazy job-quitting whiteboard girl and call it a day. Apologies for all the typos. )

I don’t watch much TV in general, but I’ve been regularly following Boston Med this season. It’s a great show, and @DrDiver and I watch it together because the source material is drawn from her day-to-day experience, which makes it even more interesting for us.

In this past week’s episode one of the patients on the show said, “I know we’re in good hands… I Googled you, and I liked what I saw.”

I bet this happens all the time these days. Who wouldn’t Google professionals before trusting them with your health, your legal affairs, or even your plumbing? So, now it’s more important than ever that a professional have good coverage on a Google search for his or her name.

There are two types of web services for professionals today: generic, broad “my first website” services and directory services.

There are a million and one sites out there for SMBs to get their first website. They’re typically WYSIWYG editors on top of lightweight templates that cover a variety of professions. I know this space well because I used to work for one of these companies. And, it’s still a very lucrative market, with a nice recurring revenue model. I think this approach is insufficient on today’s web… all the sites come out looking unprofessional, cookie-cutter, with poor SEO, and are generally underwhelming in terms of content (because that’s up to the professional to provide it all).

On the other end of the spectrum, there are directory services for professionals that also double as their web presence. These directories are places like ZocDoc, Lawyers.com, or even Yelp, which is expanding into other verticals besides restaurants. These sites hardly let the professional control the look, feel and content on the page at all.  And, often the professional is subject to crowdsourced reviews, which can be a double-edged sword if bad reviews start cropping up.

So, the startup idea is a service that sits in the middle of this spectrum.  It’s not *yet another* my first website company that broadly covers all verticals. But, it’s also not a directory. It sits in between as a highly-customizable, vertical-specific CMS for a variety of professions (I think Doctors first, and then expand to different verticals one-by-one along the way). The pages would be templatetized, so a third-party template ecosystem can blossom. And, the CMS would simplify content creation by pulling in the best reviews from other sites, or pulling in journal articles by the doctor, or other public, flattering material.

There are other ways to build on this idea, like including a online booking module, or adding on a lead gen model… but, the key is it has to be vertical-specific so a Doc has to do as little as possible in order to get a site that’s good-looking, content-rich, and well-SEO’d. The core problem being solved by this idea is Docs need to has some presence on the Google SERP for their name.

If you (or someone you know) is working on this idea already, drop a line in the comments.

Every Computer Science 101 class should open in the first lecture with this quote from Zed Shaw’s “Learn Python the Hard Way” ebook.

Programming as a profession is only moderately interesting. It can be a good job, but if you want to make about the same money and be happier you could actually just go run a fast food joint. You are much better off using code as your secret weapon in another profession.

People who can code in the world of technology companies are a dime a dozen and get no respect. People who can code in biology, medicine, government, sociology, physics, history, and mathematics are respected and can do amazing things to advance those disciplines.

I wish someone had told me this in college.  I don’t know if I would have believed this advice at the time, but looking back now, it’s so true.

I talked to a designer recently who explained to me his creative process.  The most important step for him was the point at which he asked himself: “What would Johnny do?”

I did a double-take when I heard him say that.  He was referring to Jonathan Ive (SVP Industrial Design at Apple… he designed everything you love from there) whom we had been talking about earlier in the conversation.

Since WWJD is already taken, I propose the slightly elongated WWJID. Here is the corresponding logo, ideally for tiling as your desktop background.

Marketplaces have some really favorable aspects that make them great businesses to operate. But, many marketplace businesses have a common problem they all must solve: participants circumventing the market to complete transactions.  

Buyers and sellers in a marketplace are looking to transact with each other, and the economic value that a market owner extracts is often not aligned with the interests of the market participants. Transaction fees are a friction to liquidity, a tax on the system.

Based on my experience with buying on AirBnB, they are facing circumvention challenges.  The supply side of that market (the property owner) is incentivized to use the market solely as lead generation for renters, and then take the transaction off the system in order to avoid transaction fees. This problem is not unique to AirBnB; all marketplaces face these types of difficulty.

So, what is the solution to circumvention without giving up on generating revenue from the market? It depends entirely on the characteristics of the market.

* If the market is heavy on the supply-side, you can charge a listing fee to sellers. A listing fee can act as a quality filter to ensure that listings in the market are legitamite. For example, Craigslist had a terrible spam problem in popular metropolitan areas in both the jobs and real estate verticals.  By charging a listing fee, they put a quality filter on listings while also generating revenue. Win/win.

* If the market is heavy on the demand-side, you can charge a sourcing/purchase fee. HR recruiters employ this model in the market for very high quality talent, where there is limited supply (say, rockstar RoR devs) and excess demand (say, rails startups). When a recruiter sources a new hire to an employer, they extract economic value in the form of a % of first year salary from the employer (demand-side) of the market.

* If the market is a highly considered purchase that is not likely to happen on the platform, then a lead generation fee is an appropriate solution.  For example, there are a number of marketplaces in the global B2B trade world (like Alibaba) that charge suppliers a fee for warm leads to new interested buyers.  These types of markets often intentionally hide of obfuscate necessary contact information that would allow interested buyers to circumvent the platform and access the suppliers directly.

* If the market is in a vertical where the transaction data is very valuable to a third-party, then you can operate the market for free (or cheaper) in order to reduce the friction on transactions, and then sell the transaction data to a third-party in order to generate revenue from the market. For example, stock exchanges sell priority access to bid/ask information to high speed traders for a high premium. Transactions on the exchanges are not free, but I suspect the incremental revenue from selling transaction data allows exchanges to charge less for transactions than they might otherwise. (or maybe it’s just gravy).

* If the market is in a vertical that is rife with fraud, then the market owner can charge transaction fees to either side of the marketplace participants with the value-add of insuring against fraud. For example, ticket sales are a potential nightmare because tickets can be counterfeited, lost, or delivered late. StubHub charges a transaction fee with a value proposition to buyers by insuring that they will receive legitimate tickets on time and a value proposition to suppliers that the buyers payments will not bounce.  This type of market gets more into the middle of the transaction, like an escrow agent.

I love thinking about this stuff, so let me know what other revenue model ideas I’m missing.

When generating revenue from a marketplace, it’s important to think about how revenue generation efforts will affect the market participants.  Look for opportunities when charges for services can increase the customer experience and minimize additional friction on the transactions because the health of any marketplace is directly tied to the volume of transactions.

Speaking strictly anecdotally, it feels like iPhone developers seldom open source the apps they develop. I’ve googled around and found some blog posts aggregating open source iPhone apps, but these apps are generally few and far between. Especially in the top apps released, it feels like pretty much none of them are open source.

Is this a cultural change of a newer generation of developers? Or perhaps this is a side effect of how simple it is to generate revenue from an app in the app store?

I wish I had data to support this thesis. It would be great if the app store published code license metadata, but it doesn’t look like that’s the case from my exploration.

It’s not necessarily a bad thing if it’s true that iPhone apps are more often closed source than their counterparts on other platforms, but I like the ethos of open source software and am disappointed if the app store is setup in such a way to discourage sharing source code.

The Singularity Summit is coming up in a little over a week.  I wish I could go out of intellectual curiosity, but I can’t think of any reasonable way to justify the trip otherwise… Spark isn’t investing in SkyNet just yet ;)

At the pace at which it feels like technology has been *accelerating* (not just increasing, but increasing more quickly) the past few years, I feel like I could get on board with the Singularity camp.

But, on the other hand, I just got off a plane (to Denver) and the in-flight movie was Oceans - a documentary about the beauty, power and diversity of life in the ocean. I didn’t watch more than 20 minutes of it, but it was impossible not to be sucked in my how amazingly beautiful and complex nature is. In that light it’s nearly impossible to believe that we could create a machine that could eclipse the intelligence that is the product of billions of years of natural progress through Darwinism. When viewed through the lens of appreciation for the power of nature, the idea of the Singularity just seems very arrogant about humans’ power to create technology some much better than what can emerge out of nature.

So, I’m on the fence, which is a pretty useless place to be.  I hope I get conviction one way or another soon… which is why I’m writing this blog post.  I’m sure some commenter will sway me.

And, if nothing else, you have to give Singularity believers credit for having belief. It’s tough to go out on a limb and make a big, controversial bet about the future.

I’ll close with a slightly related tangent… a quote by Eric Schmit (found via Dave Morin and Mike Galpert) from an interview he gave yesterday.  I wonder where he stands on the Singularity issue:

There was 5 exabytes of information created between the dawn of civilization through 2003,” Schmidt said, “but that much information is now created every 2 days, and the pace is increasing…People aren’t ready for the technology revolution that’s going to happen to them. [emphasis added]

The Gmail team recently launched a feature that I’ve been waiting awhile for: you can now drag and drop attachments between a Gmail compose window and the desktop (and back again).

It’s these small usability problems (which have been possible in a native-app based mail client since 1998) that hold Gmail back from greatness, and I’m glad to see them slowly get solved.

But, in the case of this feature (and I imagine others in the future) the feature is only supported in Google Chrome.

Google is starting down a path of vertical integration that I think will ultimately hurt the end consumer. Google supported a diversity of browser choices for years, by helping Firefox financially and contributing code to other open source browser rendering engines and projects… until of course they launched their own browser.

Not everyone will use Chrome, and Google should be focused on giving everyone the best possible experience on every browser.

I was talking with Bijan the other day, and he mentioned he heard it’s possible to get through a Computer Science (CS) major these days without learning the language C.  I had heard the same thing (though I couldn’t tell you at what program that’s true), and I do know for certain that many introductory programming classes have moved away from C in favor of Java.

It’s really surprising that you could escape an undergrad program with a CS major without touching such a core tool as C. If you don’t learn C, how will you know the origin of Hello World? Of course I say that tongue in cheek, but I do feel that real-world engineering will require strong C skills at some point or another, missing out on it in undergrad would be a shame.

I don’t know exactly what an undergrad program in CS looks like today, but I’m a pretty young guy, so I’m going to assume it looks pretty similar to what it looked like back when I was in undergrad (minus this C controversy).

It’s remarkable how little a computer science education matches up to the real-world building of large scale applications.  I think the whole disipline would be best served if it were broken down into a few component parts, rather than being the hodge-podge that it is today.

Paul Graham poked at this subject in Hackers and Painters, so big hat tip to him for getting my mind going on this subject. If I had my way, I’d explode CS into the following pieces. Each might be its own major:

Architecture: This is what I think an academic CS prof spends most of their time thinking about… building new languages, operating systems and compilers that will probably never see the light of day in real-world engineering… but, architecture is an important discipline in that all CS professionals are standing on the shoulders of architecture giants that designed the tools used across the discipline. This would be the most academic track.

Networking:  The networks we rely on have gotten so complex, with so many layers of abstractions in the lower levels of the stack, that I feel this could be a discipline of its own at this point. This component of CS would likely be an academic/engineering hybrid: some network theory, topology, etc… and some flavor of modern day sys admin skill set on top of major networks at scale.

Engineering: This is how to build something massive. It some ways it’s the opposite of the architecture track… for example, if you take a Database class in CS undergrad, you will be taught how to create am ideally normalized database where data is only found in one table and can be joined with data from other tables via keys. In the real world, major databases at scale are highly denormalized, because joins are too computationally expensive, and so data is duplicated everywhere, but everything is optimized for speed. 

This real world example of databases is how I think about engineering. It’s using what you know empirically about performance and speed to implement what an architect or designer wants to do in the most efficient and sustainable way possible, even if it would make academics and architecture nuts roll in their graves. In a typical CS undergrad program, I think engineering is the biggest hole in fresh CS graduate’s skill set.  They typically pick up these skills in a summer internship where they see real-world code being developed at a big company.

Design: Software design (the UX/UI side of design) is closer to sociology than any other discipline, but it’s in the hodge podge along with many other course in a CS major. Understanding how to do needs finding, contextual inquiries, iterative design with user feedback, and even the art necessary to execute well-intended designs… these types of skills would all live in this track. Some school like Carnegie Melon and Georgia Tech have independent programs in HCI, which is a step in the right direction, but this is typically a subset of a CS major, and yet there is very little traditional (think Architecture when I say “traditional”) CS involved.

Applied Math: Computer software is a finite state machine, and as such, all of CS is really just Discrete Math. It’s not even all of Discrete Math… it’s actually just Applied Discrete Math, because Discrete Math can be done in theory, on paper, which some CS majors are forced to do in a class or two, but this process has little real-world relevance.  That said, CS departments are full of guys pushing the envelope on Applied Math, and if someone wanted to major in just that discipline, there’s definitely more than enough meat on the bone to do so.

So, that’s how I’d split it up. What did I miss? I was trying to figure out where key tools like AI, Machine Learning, NLP, etc fit in… but in the end I decided that these topics are not organizing principles… they’re applications of the breakdown I prescribed above. Feel free to tear this apart in the comments here or on your own blog.

As computers are now an essential tool of every type of engineering, it seems silly that we have a department that could grant you an undergrad degree in computer science.  It’s like granting you a degree in Abacas Science 100 years ago (another finite state machine)… no one would do such a thing… so why do we do it today?

Disclosure: I wasn’t a CS major in undergrad, so maybe I have no idea what I’m talking about, but I did take courses all over the department and saw first hand what a hodge-podge it was.

Sometimes I ask people I know in my life who I believe would be good entrepreneurs, “have you thought about doing a startup?” I common response to this question is, “Yes, but I don’t feel like I have a good idea.”

I typically let the conversation end right there, because I would never push someone to do a startup that is hesitant.  A good entrepreneur has an undeniable drive, a hunger to bust through brick walls and make their startup successful. It’s not possible to try to convince someone with logic to adopt that often-irrational entrepreneurial frame of mind.

But, if I were to continue the conversation, I don’t think a lack of a good idea is a reason to not do a startup. Want a good idea? Here’s 30 great ones that Paul Graham wants to fund. Why would Paul Graham just give away 30 great ideas for free? Because the value of a business is not in the idea. In an early stage company, the value is in the people involved.

It’s the same reason that anytime I have a startup idea, I immediately write a blog post about it here.  I know that ideas are cheap, and additionally, ideas get better as people contribute feedback to them, shape them, and make them their own. I’d love to see a great team make any of the nagging startup ideas banging around in the back of my head a reality.

Entrepreneurs almost never take a startup from idea to full execution in a straight line.  The idea is a just a starting point from which people weave and bob around as they try to figure out where the money is in a new market they are creating, or existing market they are co-opting.

The converse is true here too.  If you’re really passionate about doing a startup because you think you have a great idea, you are actually starting with very little. In fact, the passion you have is very likely far more valuable than the idea itself. So, I would never discourage someone in that position, but my advice at that stage is typically: go get the 2 or 3 smartest people you know and run with the idea. Because people are so important, it’s really difficult to give the idea a fair shot if you do it alone.