softdev12

Whenever I see charts that focus on statistics like IPO valuations (e.g. #4. High-end IPO valuations are rising dramatically), it makes me think that people have forgotten the cost differential in a starting a software-based business now vs. 1995. When people talk of the few (<20) winners being multiple billions of dollars, they don't consider all the thousands upon thousands of tiny sub-five person startups that quickly try an idea and fail. So the total number of startups in the "universe" of startups is probably an order of magnitude greater today than in 1995 or 2000. It makes sense that more winners would come out of a much larger pool of companies.

I wish people would flip the argument and instead of saying tech bubble, start asking about the death rate of all the millions of apps, small shops, and side projects that are flooding the tech sector. I would like to see if this "fail" rate is too low (in the sense of not having enough winners) rather than a tech "bubble" being too high.

While I appreciate what you are saying, I think this statement is not really accurate if you look at the facts. I would not say "virtually any" because there are numerous academic projects that are now abandoned and whose creators consider a failure. So, semantically, "some" might be a better word in the place of "virtually any". In any case, I think you are underestimating how complex Airbnb's tech is. They have tons of banking regulations that they have to follow because of their payments and to think that their tech is "crud" is just a testament to how well they make the complex seem easy.

I enjoyed reading this because it's not something you usually see publicized. And as a single data point leads me to believe that there are some people really doing well with these sort of tactics - just not talking about. It's like the blackjack card counters in 1980's and 90's. They were successfully employing a system to pull money from casinos - but weren't talking about it. It was only when the MIT Team started publicly talking about the strategy, that it made it into the open. And that was at the time that the casinos made it not a really viable strategy.

So it seems likely that there are successful strategies for using these affiliate tactics for profit. I know that Wired magazine profiles a few of these people every once in a while. There was a Canadian entrepreneur who was big into the diet pill scheme and made millions (he, too, would fly down to Vegas with friends). And here is a link to another scheme that used scareware:

http://www.wired.com/2011/09/mf_scareware/

This really surprised me. As a developer, I loved being able to iterate new app versions in under a few hours - live to Google Play. In fact, it was a major factor in prioritizing Android over iOS. The Apple review process, in my opinion, is overly draconian.

I had expected Apple to become more like Google in their review process - more friendly to developers. But it seems the reverse has happened - and Google is becoming more like Apple.

This suggests two things: 1) there are just too many apps now and discovery is too hard (So Google is trying to put the brakes on the app flood through a review process); 2) there are too many bad actors - so all good actors are being screened extra carefully.

The algorithm seems to run on the following factors per the article:

"job tenure, geography, performance reviews, employee surveys, communication patterns and even personality tests to identify flight risks".

I wonder if employees can try to game the algorithm by artificially ranking high on their score for leaving - with the sole purpose being to get a pay raise.

I really like the idea. Testing on ios is beyond a major pain. When I read the title to the post, I thought you found some way to install on a real device without having to go through the testflight process (and submit the app to apple for beta review) via a simple install link. But it seems that the tester needs to have Xcode (i.e. be a developer caliber installer). This doesn't take away from the idea, though. Any improvement in ease of testing is huge.

I previously used testflight before Apple acquired it. They recently shut down the standalone website. I found that process way easier than the current process. All I had to do was send an email to a non-technical user and the app was installed. Now it's more hoops to jump through and ladders to climb.

This is interesting because it's the first time I've heard about 21 and they have raised $116 million pre-launch. Given that companies who've raised an order of magnitude less are all over the news, it's surprising they have been so stealth.

I'm also still not clear what 21 does. According to the article they are building out the blockchain into the internet of things (IoT). That sounds really vague. Plus, the article mentions a contractual twist - like smart lawyer contracts. These two things don't seem to obviously go together.

I read a bunch of different tech sites, but it my opinion, the most relevant stuff is on HN. It basically aggregates all the best stuff from all the tech publications anyway, so you are probably just as well off in the end. Of course, if you want to read more academic articles per se, stick to the big science publications like Nature, etc.

Here's a list in no particular order:

1) recode

2) pando daily

3) techcrunch

4) techmeme (another good aggregator)

5) Mit tech review

6) New York Times Bits

7) WSJ D

8) CNET

9) Mashable

I now use a quick test whenever I hear about a new startup: I ask myself what's the "trick" they are proposing. If the trick is sufficiently clever, then I raise the odds of the startup getting initial traction.

Here are a few examples:

1) rent out cars that people leave at airport parking (i.e. flightcar, etc.) 2) group people together for a buying discount (i.e. groupon, etc.) 3) a middleman service that uses middlemen (i.e. magic, etc.)

Most of the hot new startups aren't delivering on a basic need (like food or shelter), but seem to be like one-off computer hacks. And that's not a bad thing. These are smart people bringing out clever solutions that deserve talented valley employees.

The returns that these startups will generate are much more attractive for short-term investors and will have clear advantages over the long-term mentality needed for building a hard new technology.

It was a good read. The most interesting parts were the backstory behind how MySpace got started. In particular, how the company pivoted a number of times before getting the social network idea to really catch on. It was fascinating to see how they went from selling these cheap toys at kiosks in the mall to becoming a tech media network.

I'm just finishing reading another book called Click by Bill Tancer written around the same period. And Myspace had 3 out of the top 10 searches for brand websites (myspace, myspace.com, and www.myspace.com). Facebook had just appeared on the scene and was in the 10th search spot.

right. 7) a great team is probably even more important than luck.

I think this article has it backwards. I think that the default state for a startup is to fail. That even if you do almost everything right, you still have a high probability of failing. And that there should be a list of all the things that need to be done right in order to succeed, rather than listing all the reasons for failure. For example:

1) make a product that people want (vs. no market need) 2) attract investors (vs. ran out of cash) 3) media virality (vs. poor marketing) 4) perfect timing (vs. product mis-timed) 5) have clear distribution to market 6) lots of luck

#6 is not mentioned at all in the article and is probably one of the biggest factors.

Summary of this article:

1) Barry Silbert's BIT has beat the Winklevoss' (or Winklevii) bitcoin fund to market by using a backdoor listing technique.

2) Finra granted its request for a permanent ticker symbol, GBTC, and that “is expected to be effective shortly.”

3) Each share of BIT is worth approximately one-tenth of a bitcoin. As of Friday, the trust’s net asset value stood at $24.43 per share.

This is awesome. When I read the previous post by OP that hit the number one spot (called Consulting), I immediately wished that there was a post-mortem on the failed startup.

This is important for many reasons, but the least of which is that there's too little written on what goes wrong with a startup. Everyone (especially the media) seems to think that every startup becomes one of the billion dollar darlings that saturate the news.

My take on the overall concept was that it was going to be really hard to pull off, because you were building a two sided marketplace. And that's exactly what airbnb is (which is why it's appropriate you called the startup an airbnb for cameras). Even if you had you perfect execution and built a product that was 2x better than anything that existed, getting a two-sided marketplace to work requires an extraordinary amount of luck, marketing, and timing.

The flipside is that once a two-sided marketplace exists, it's really tough to displace (for exactly the reasons that it's so hard to start). So I would have loved to hear your rationale for thinking you could overcome the odds of building a two-sided marketplace. It's twice as difficult as a normal startup because you need both the sellers and the buyers. And one is not valuable unless the other is there at the same time.

There is a weird phenomenon going on now that relates to design. i agree that the page is super simple looking and agree that this is a turnoff for most people. But for some reason people now think that design is even more important than the service. In my opinion, it's the "judging a book by it's cover phenomenon". I wonder how many people would say that the design of Hacker News is too simple?

i agree completely with your first sentence. Having distribution is just as important as the core use case of the product. In the beginning, it may actually be more important.

One of the most fascinating things about the recent explosion of the Magic service, was how Hacker News (and Product Hunt) essentially acted as a growth hack for a product that was developed in a weekend. Given that the people behind Magic were in the YC batch and probably had access to strong upvoting on Hacker News, that seems like a really good example of a growth hack (where the product is the marketing).

This is interesting and seems to indicate that there is a small edge case for people who want as much privacy as possible.

The larger topic for me is that webRTC has to make tradeoff for developers. And, I for one, love what it does. If you look at legacy comms development in the peer to peer space (e.g. Skype) the process was orders of magnitude more difficult than using a webRTC implementation. So, as a developer, sacrificing a prompt that grants an IP detection notice seems like a worthwhile tradeoff.

Also, I think webRTC is still in draft form and is still being modified in working sessions (although I'm not 100% sure of this). It would be great if Apple would get on board with webRTC for ios.

I agree. For the vast majority of use cases, this doesn't seem to be an exceptionally big issue. Almost all major websites collect IP addresses and don't explicitly prompt the user that this is happening. It seems the OP is really highlighting the edge case for users who want quasi-absolute security.

I couldn't agree more with your full post. It's almost become so obvious and cliche that people have forgotten why business exists. It's as if all the media hype puts the focus on the riches of the founder and not the value created by the product.

Every once in a while you see a publication do a retrospective of a failed startup. And every once in a while you see a blog post from a startup that exploded. But the statistics are completely skewed with distorting the reality of startup success. Mentioning that there are now X number of billion dollar startups masks how many startups are simultaneously failing. I would love to see a study that showed the total number of inflation-adjusted billion dollar startups as a percentage of the total number of startups started over time. This percentage may actually be decreasing, even though the media seems to be suggesting that there's a tech bubble.

I just got 2 emails today from the legacy testflightapp.com, reminding me that they are shutting down the site on Feb 22 and that I no longer will be able to access all my historical tester-app data. The email provides instructions for migrating the data from testflightapp.com to itunes connect (which is always a pain to have to go through). When I went through the steps in the email, it seems as if all the previous accepted testers would have to re-accept the testflight process. I hope this isn't the case, but it's not the end of the world otherwise.

On a larger theme, I'm surprised that someone (an enterprising tech person) doesn't set up a service that archives all the data from startups that get acquired and then are shutdown. I'm imagining something like a secure github that has all the channels for each dead/acquired service that you could log into. Or perhaps there could be a company that charged the acquirer a small fee to take over the hosting of just the archived data, so it's still accessible. Perhaps this already exists?

I'm glad that the article highlighted how valuable the mapping information is that Google collects. While Uber probably could still run its current operations without Google map data, there is a much smaller probability that they would be able to run a self-driving car service without a highly accurate mapping product. And to build a mapping product is really difficult and expensive. I was surprised when Google starting putting money into mapping the streets because it's a giant effort with major barriers to entry. Nokia bought a company called Navteq that was the clear leader (and basically a monopolist) in driving the streets and collecting the data (they originally supplied data to Google Maps). Navteq supplemented the data with satellite images and other sources, but the roads change often enough that you need people driving on the streets.

So, Uber really needs to start thinking about the mapping aspect. Perhaps they can incent their current drivers to start reporting back all the street info as they pick up their passengers. Of course, this would bring up problems if the drivers eventually realize that this data could be used to replace them. It would be interesting to see uber cars with Google StreetView style cameras on their hoods.

I guess the problem with growth hacks is that once they are known, they get overexploited and then lose all their effectiveness. Also, it would be nice to not have to sign up to an email to receive the hacks. Perhaps, have a few hacks that you list for free on the website and then a login to receive new ones.

Good luck with this. It would be interesting to see if you can use your proprietary growth hacks to make growthup into a huge site.

I know that NoSQL has a bunch of advantages with super large datasets, but MySQL really is a solid technology and is underrated in how scalable and robust it is.

You'd be surprised how much MySQL can do and how great the documentation is.

The article just skims over this part:

"To start off with, Goodfellow and co place some limits on the task at hand to keep it as simple as possible. For example, they assume that the building number has already been spotted and the image cropped so that the number is at least one-third the width of the resulting frame. They also assume that the number is no more than five digits long, a reasonable assumption in most parts of the world."

This seems like a huge task. Someone has to go through all the thousands of images and first crop them? During that time, it would seem like they could just input the number into a database.

Maybe I'm missing something, but I read the "cracked" part to be a totally automated system that scans all the pictures and pulls the numbers with no human manipulation.

The general sentiment on this thread pretty much sums up the idea of "scraping as a service". To me, there is definitely a legitimate business need to be able scrape. Whether people realize it or not, companies have in-house teams that build custom scraping tools. The challenge for you is going to be able to siphon out the bad actors that may use your service to do things that you would not approve of.

i've read all his work, too.

I enjoyed the biography because it gave context to his interviews. He would say things in interviews that were clearly misleading because of certain things he had gone through in his past. The book highlights the reasons for why he would say these things. And if you go back and listen to his interviews you get a new take on things. It's as if you are hearing a new interview (after reading Ghost Story).

I would say that while his biography is worthwhile, it's obviously less worthwhile than his writing. And, in the end, I preferred the "Although Of Course" bio-book a bit more than Ghost Story.

I just finished Virtually Human and would definitely recommend it. It makes a very in depth analysis for why "conscious" computers of the future will eventually want human rights - the same way that certain minority groups fought for equal rights throughout history.

The book seems to argue that the likely future scenario is for humans to replicate themselves in a separate computer entity. So people would have a duplicate robot twin. Very little mention is left to addressing the case of people actually merging with computers (so one entity and not two). I think this case is far more likely.

To really hit the mainstream and get huge, VR needs to address the "hype" build-up in new technologies. What I find singularly unique to VR over the last few years is the announcement of one breakthrough company leapfrogging another with just spec details, videos, and big funding. First, there was tons of excitement with Oculus with Carmack joining. But then Magic Leap seems to have developed something even better. Perhaps something new is about to be announced beyond Magic Leap (only in a video demo of course). It makes the mainstream consumer not want to purchase with the fear that something new is going to put the hundred dollars of payment to waste.

Perhaps, VR companies would find the Apple secrecy approach valid. Only releasing real products that are ready to be shipped in mass production within a few months timeframe.

I think the key sentence in this post is where they mention launching the MVP with 20 companies. If these are all paying, large businesses that seems like the founders had deep relationships already established in business. For most startups, launching with 20 businesses out of the gates would be a huge win.

In fact, one could argue that the real value of accelerators isn't their mentors but the fact that they provide an immediate source of other businesses willing to try (and pay) for the product. So for a company in an accelerator, it seems much easier to get to 10,000 paying customers than for a company that is building a product in a garage (without an accelerator).

In my opinion, going from 0 to 20 paying businesses seems to be relatively harder than going from 20 paying businesses to 10,000 total customers. Meaning that if you get traction with 20 businesses that probably means you're touching 2000 people indirectly (assuming 100 people companies) already.

looks good. nice to see that people are still trying to rethink objects that have been around for years.