Build What Had Previously Not Been Possible | Hacker News Reader

Build What Had Previously Not Been Possible

jasonlbaptiste.com

164 pointsjasonlbaptiste64 comments

From "You and Your Research" by Richard Hamming (discussed at http://news.ycombinator.com/item?id=229067) :

    Most great scientists know many important problems.
    They have something between 10 and 20 important
    problems for which they are looking for an attack.
    And when they see a new idea come up, one hears them
    say "Well that bears on this problem." They drop all
    the other things and get after it.

    ...

    The great scientists, when an opportunity opens up,
    get after it and they pursue it. They drop all other
    things. They get rid of other things and they get
    after an idea because they had already thought the
    thing through. Their minds are prepared; they see
    the opportunity and they go after it. Now of course
    lots of times it doesn't work out, but you don't have
    to hit many of them to do some great science. It's
    kind of easy. One of the chief tricks is to live a
    long time!

To accomplish this, it helps to keep a list of important problems that haven't been solved yet. As you hear about incremental improvements, it may suddenly be within reach, when popular knowledge says otherwise. Online video was obviously going to be a killer app: one Victoria's Secret Live fashion show brought the Internet to its knees several years before Youtube existed. There just wasn't a way to do it well before the cards fell into place: Flash video cut through the Gordian Knot of browser video plugins, and low-cost server infrastructures finally became cheap enough to store petabytes of video. The casual observer may miss the events, and many may jump in too early, but it helps to know what problems need to be solved.

What sort of problems are people thinking of?

SSDs have been a big speed improvement - have they enabled new developments?

RethinkDB thinks so.

In mobile robotics you can record data on the move without worrying about damaging the disk or enduring sudden massive write latencies. So you get more interesting datasets using SSDs than you could with traditional disks.

Good point. Maybe jasonlbaptiste's next article would be a list BIG problems still waiting for good solutions. Here are a few I can think of:

- Email 2.0 (Wave was a good experiment but we need more)

- Online Dating (still broken)

- Online job search (could be improved a lot)

- Search 2.0: A Knowledge Engine that understands the question and finds you the answer.

I don't believe this theory of 'build what had previously not been possible' because it is not true and does not match evidence.

An example is the readiness of markets. Many successful web 2.0 companies were re-implementations of old dot-com ideas, with the technology being the same (bar frameworks that minimized development time) but with a more ready user market and refined implementations.

For example: XDrive - Dropbox, Friendster, theGlobe.com etc. - Facebook, Yahoo bookmarks - delicious, boo.com - etsy, etc., webvan - amazon delivery, etc. etc.

Further, many of the technologies cited are old (GPS is almost 40 years old). It is economies of scale that resulting from market acceptance that make these technologies more viable. Most software products do not require bleeding edge technology, which is often a curse (ie. using HTML5 today when 80% of browsers do not support it).

If you think about what made YouTube, Facebook et al successful, it wasn't some magic new technology, it was the right implementation and market acceptance.

This bit has been done to death, but what is different today is:

* It is faster to develop apps because developer time is more efficient and scalable because of tools, but there is nothing in RoR, Django etc. that couldn't have been done 5 years ago with PHP (and I mean that in the end result, not how it is done).

* Broadband penetration (especially globally)

* Browser penetration (IE5 introduced xmlhttprequest in 98, but it took 7 years to reach a decent use base for Ajax apps to be viable)

If you went by the advice here and intoduced a bleeding-edge Ajax app that was bandwidth heave would have failed in 2000 (and they did, we would use xmlhttprequest for intranets only).

The risk with this advice is entrepreneurs betting on new over better and becoming complacent in believing that what they have as better and 'previously not possible' should win.

(edit: so I guess you could say 'build what had previously not gained market but would today because of tech penetration')

I think all he meant you don't have to invent something brand new, but you said you should take advantage of the new tools, technologies, social graphs, etc. that were not previously available. It sounds obvious, and maybe it is, but it's a good reminder and he put together a great list with great examples. Hat tip / thank you to Jason.

> If you think about what made YouTube, Facebook et alsuccessful, > it wasn't some magic new technology, > it was the right implementation and market acceptance.

I would have put that as the right implementation at the right time

many implementations will fail simply because of being too much ahead of time

the "right time" imho is not only about market acceptance or market penetration it's also about "when people are ready to accept it".

Before google decided to use xmlhttprequest for gmail, nobody would want to use it because it was considered "IE only", but once google used it, it gave xmlhttprequest a certain credibility and then suddenly everyone started to use it and then the word "Ajax" was created.

In another context, take smartphone publishing their location in foursquare or whatever, it's not only because they have the technology to do it (GPS), it's also related to social acceptance, after people got used to display their personnal infos on facebook, myspace, etc. then it is just a little jump to accept to display their geolocation.

I do believe in "build what had previously not been possible" but it's not only because of technology or market penetration, there is also a big "social acceptance at the right time" parameter.

You could certainly build things that were possible before, but I'd rather suggest entrepreneurs go towards that, which wasn't. And if you do build something that was possible before, ie- Dropbox, then I suggest you use the building blocks to build an iteration that wasn't possible before. Dropbox was started as a small project by Drew and eventually a YC funded company with a small 25k in capital. It wouldn't have been possible to do Dropbox before AWS,etc. without spending a ton of up front Capex in hardware you may or may not use down the road.

One of the first genuinely hopeful bits of entrepreneurial thinking I've clicked to on Hackernews in quite a while, something that applies only to people who are thinking up highly original ideas likely to create lots of new economic value, and not to people doing the same stuff everyone else is doing in hopes of a quick cashout.

Generally good advice - although I would say that there are things that haven't been built today that could have been built yesterday and that could be the building blocks of tomorrow. Facebook and YouTube are two examples that weren't built using state of the art technology, that could have been built a few years earlier. I think their success was in their execution not the conception of the ideas as they weren't actually original ideas but well polished ones.

YouTube is actually interesting. They attribute their rise to a few building blocks: http://www.youtube.com/watch?v=nssfmTo7SZg

I wouldn't put all of these as some sort of future stack building blocks, but there were enough things there that made something possible, that wasn't possible before: easy video sharing.

1- broadband in the home 2- emergence of flash, so no codecs required 3- proliferation of digital cameras 4- cheap hosting 5- one click upload 6- ability to share embed

I'll just leave this here. http://bit.ly/cOGT5n

Full link: http://www.youtube.com/v/kfEzHdWKOoQ

Also worth considering: Build what is possible today in US but not in Europe/Asia. i.e. Watch what your fellow US entrepreneurs are working on and adopt it to your local.

Very true and very lucrative. Only pitfall is making sure you can translate across cultural norms. CyWorld tried the opposite Korea => US and failed.

There are any number of opportunities to bring e.g. Japan into the 21st century. By the by, this is true for lots of folks demographically dissimilar to this forum in the US. JavaScript is cutting edge in my market, no joke.

Why would we want to bring Japan back from the 22nd century to this mess?

Can you elaborate please?

Do you advise this from the perspective of being (1) in US or (2) in Europe/Asia? It doesn't make sense either way, and people from other places might feel excluded anyway.

From a European/Asian perspective. Example: Barcode scanner/price comparer iPhone app. Although we have many local price comparison services; there is no such app to read (most of) the barcodes where I live.

    > people from other places might feel excluded anyway.

I apologize for narrowing this down to Europe/Asia, this should apply everywhere at least a couple months behind US in adopting technology.

I like the headline idea, but still wonder when you say:

: If 9/11 happened in 2010, how much different would it have been? It took forever to know what was going on as the channels of TV and Radio weren’t fast enough. The web wasn’t ubiquitous and the ability to publish as fast as today was not there

I was watching it on TV and hanging around on the Ars Technica forum, but I can't really imagine my life would have been much different if I hadn't heard about it for a month.

Similarly, location based services - there seems a huge limit to what kinds of software could be much improved by being location aware.

Is this me being unimaginative or are you just picking interesting sounding things and saying they must be futurestack components because they exist?

Well a few things:

* The tragedy would have been the same. By no means (and i don't think you are implying that I am either) saying tech today would have prevented it. * The zero delay of information flow would have made things a lot less hectic for everyone. At the time the towers fell there was a massive panic of finding where everyone was and if they were okay. I could imagine a popular hashtag emerging such as: #imokay . It wasn't possible to broadcast to all of your friends in a matter of seconds if you were okay in 2001.

*I think we've scratched the surface of how locational context can improve most things. If we can organize and provide context to information in the real world, the same way we have to the digital world, that will be big.

You could imagine twitter not crashing and burning itself after an event like that?

Theoretically you would hope it stays up, but past data shows that is iffy. FB might be sluggish, but it would be fine.

The web was still pretty grown up in 2001 - admittedly no facebook or twitter; the idea of immediate spread of information through non-traditional routes was not widespread.

I checked cnn and bbc news for information (both of which went down several times that day I might add)

But I really doubt twitter would stay up if a 9/11-scale event happened today. Even facebook I'm not sure about. If the CNN site couldn't handle the traffic I can only imagine what that spike looked like.

"It wasn't possible to broadcast to all of your friends in a matter of seconds if you were okay in 2001."

well, there was - cell phones and SMS. Those networks went down too (source: http://www.continuitycentral.com/feature0153.htm )

SMS was certainly there, but nowhere near the usage that it has today.

Hmm, apparently the UK were 4 years ahead of the US in terms of SMS adoption (source: http://www.intomobile.com/2007/07/23/americans-are-4-years-b...) so maybe you're right. Certainly felt to me like SMS was well established in the UK in the 2000s.

SMS and phone calls don't broadcast, that is the key difference.

It wasn't possible to broadcast to all of your friends in a matter of seconds if you were okay in 2001.

I think you're vastly overestimating the true popularity of twitter. If I where to tweet #imokay very few would see it, and those who did would have no idea how to contact my family to let them know that I was in fact OK. If I told many of my friends to "tweet the hashtag #imokay" they'd just look at me and only be vaguely aware of what I was talking about. Even in 2010 the best and quickest way for most people to let people know that they are OK is to make a phone call.

who says im talking about twitter only? I actually think it's not as vastly popular as people estimate too. In the instance of 9/11 the best way to let a large number of people know you are okay would certainly be Facebook/Twitter/etc. Phone calls would take way too long. You should obviously save that for close friends.

Uhh... When I came home and saw the second plane hit on TV, it took about 10 seconds before I and most of my friends were on IRC, on a moderated channel that got realtime news relays. Actually, there were several such channels, since they hit capacity pretty fast. That's how I got all the valuable information, like how there were reports that a plane was headed for the Eiffel Tower, that I could share with my offline family.

If I had wanted to let someone know I was okay, I probably would have tried SMS, found it didn't work, and then made a note of letting some blabbermouth know once I found a landline or netcafé. 'course, I didn't live in the US, so this is pretty hypothetical.

The point is, I got news (and could discuss with friends) faster back then than I probably would today.

I was six blocks north of the Towers when they fell. The reason it was hard to get the word out had nothing to do with the presence or absence of Twitter / FB / etc. The reason it was hard was because all the cell towers were down (literally down: many of them were on top of the Towers) and the landlines were clogged with several million people making or receiving calls.

Were this to happen in 2010, you'd see the exact same results. It might be a bit better, because more people would be using VOIP and the land-based Internet didn't go down, but that wouldn't have solved the issue.

Augmented reality (and similar vision-AI applications) deserves a mention here. It brings together many things that weren't possible "yesterday" - fast mobile processing, location awareness, new sensor hardware (anything from cameras to accelerometers and gyroscopes is applicable), social/web data...

When looking at an idea it’s useful to ask yourself: “Would it have been possible to build this company 12-18 months ago?”

This reads like a list of stuff that has been possible for a lot longer than 12-18 months:

  Direct Access To Customers
  Anything as a Service
  Always Connected
  Zero Barrier Distribution
  Social Context
  Assume The Device Is Portable
  Location Aware...

I'd say dig a bit more, all this stuff was possible in Feb 09. You have the right approach, but the wrong examples.

I loved the article, but there is subtle problem troubling me over here.

How do you build the previously impossible without knowing the real extent to what is possible, or not possible? It's one thing to see trends, but quite another to take in depth domain specific knowledge across multiple interconnected topics and form a picture out of them. I can sit here and talk about trends in a lot of industries, but questions like sustainability, future growth require that picture. You need to get it.

It takes a lot of time to do that and there really is no way to escape it. (It's pretty awesome if you fail though, at least you learn something out of it)

However, what happens if you're in a crunch and you're trying to enter an unknown field? You can't assimilate information at a reasonably fast pace, and what you do learn will have a greater probability attached to being inaccurate/flawed. Importantly, you won't have the complete means to detect it.

I am asking this because pretty often I find that most of the overlooked problems lie in the intersections of things, and I've realized that it's really important to know how to bridge that gap before jumping in.

The building blocks he lists are things that are possible right now. It's not a matter of figuring out the future, but rather seeing technology trends and creating a unique mix of those things that were previously impossible. It doesn't have to be in an unknown field or about unfamiliar topics. Technology changes many things, as it changes, can you do something new now?

This shifts the problem from the difficult (predict the future) to the straightforward (applying building blocks). It's not an automatic business plan or a bullet-proof vest, but the unwritten assumption is that by looking for the opportunities in the changing tide, anything you create will already have a first-mover advantage by default.

Other good building blocks to use are Hadoop (map/reduce) and Lucene/Solr (search).

P.S. Great post.

Would you put Hadoop under scaleable to billions?

Yeah, Hadoop is parallelizable by design, so it's just a matter of renting out EC2 clusters. In any case, I'm not talking about competing with Cloudera, but using Hadoop in some new specific vertical. For instance, an ad network that uses Hadoop could mine user/user-clickstream data better and get more clicks. That would be using Hadoop as a building block. You could also use Hadoop to power awesome analytics on, say, an e-commerce platform.

It is off topic and my (irrelevant) personal preference, but please, pretty pretty please, find something more beneficial to humanity than convincing people to purchase things that they don't need. Innovation is exciting when it is about ending a war, or reducing pollution, or creating cheaper energy, or making it easier to find information. Every time people talk about "mining the clickstream to increase ad revenue" I have the temporary urge to abandon computing become a farmer.

kthxbye

You could also use Hadoop to analyze data from, say, a global network of distributed pollution sensors to find factories that are polluting. Anyway, the point I was trying to make is that Hadoop is a nice building block. You could use Hadoop to mine any large data set, my examples were arbitrary.

Greed has done more to improve humanity's standard of living than philanthropy ever will.

Greed has done more to improve humanity's standard of living than philanthropy ever will.

Standard of living, sure. Quality of life... not so much.

This is why the media and information markets are so compelling. Find an existing entrenched player, identify the disruptive technology they used to enter the market (often it's something like motorcycle couriers), identify the technology which disrupts that advantage...

Hey Jason (or anyone else) SMS was mentioned as part of immediate communications. Is this true yet today? A sticking point for me with SMS was that messages were randomly subject to hours of delay upon being sent because of the SMS distribution system. It's because of that along with knowing that far more robust (and free) email on devices was on the way that I never got behind SMS. However, I now have something in mind which could apply SMS, since it's available to more phone types, but it needs to be immediate. I did a check on Google, and the FAQ at Twilio as well, but can't find whether this potential lag is still there. Does anyone have insight on this?

SMS is not guaranteed to deliver, either within a certain time boundary or even at all. I can only really speak from my own limited experience on AT&T and Verizon, but those delays seem to largely be gone. I haven't had a message delay in quite a long time. If it's going to deliver at all, it'll deliver within seconds.

Yeah, I imagine the technology should have improved from a few years ago when the delays seemed prevalent. You're also right that some messages may not go through at all. I'll probably need to research deeper into the current technology and issues surrounding SMS.

Anecdotal, but I have experienced an SMS message delivery delay of about 7 hours within the past week (on Verizon). I would agree that delays are less common, but they're by no means over with.

That's what I was afraid of. Thanks for the info.

I think it's at the point of immediate, but subject to delays just like any other technology. I haven't had an SMS delay in quite a while.

I would curious to see more insight into the history of SMS, technology behind it,etc. Wikipedia does an okay job. Correct me if I'm wrong, but SMS does not use the data network? I have been able to send sms messages with no data service before.

Yes, from when I looked into it the bottleneck seemed to be the SMSC (short message service center), which receives the message then sends it out to users when they are available. Apparently, SMSC's often became overloaded with the popularity of SMS. You're right, SMS I believe takes place on a separate mobile network rather than the open Internet, which is why it's subject to the fees and different behavior. I'm at a point where I'm torn between writing SMS technology off in favor of email, or considering it for certain things, because we're still a long way away from everyone comfortably accessing their email via mobile.

"Transactions <= .99 Microtransactions used to be a pain in the ass."

I would argue they still are–especially on the mobile web. I was talking about this with a co-worker ... unless you want to give Apple 30% of an in-app purchase, it seems there's no easy way to sell an item or subscription on your phone. People don't want to enter their billing information–especially on a 3" screen.

Until we have some sort of secure/universal/mobile "ONE-CLICK" purchase solution that doesn't include usurious rates, microtransactions will continue to be a pain.

Facebook and Amazon are the only ones I could see pulling it off–though maybe it shouldn't be a single corporation doing it.

Are you aware of PayPal's Micropayments feature? $0.05 plus 5%, meaning a $0.99 purchase only costs 9 cents (they round down).

https://micropayments.paypal-labs.com/

Note: this is the third time I've posted this link on HN in about 5 months, but each time people have thanked me for it so at risk of looking like a shill, here we go again...

Yes, I'm aware of it ... but it's not the solution I'm seeking.

Off topic, but I absolutely cannot stand the way that this site renders and behaves on an iPad. What's wrong with it? It wobbles all over the place and the text looks non-native and jagged.

Sort it out, Baptiste!

yikes, not good. it's being done through PadPressed.com , app we released last week. Just put out a new version and it's using a new cached plugin. Can you please email me a screenshot: j@jasonlbaptiste.com Thanks!

I'm assuming you have an ipad. The problem occurs in landscape mode. Try scrolling and you'll see what I mean.

Correct. Looking at it now and it looks okay and doesn't wobble. Weird.

Here's how to replicate it on my iPad at least

In landscape mode, put your thumb on the screen and hold it there. Then move it slightly to the left and stop. The page continued shifting despite the fact that I'm not dragging anymore and eventually it just shunts itself off the screen entirely.

I noticed this too. Seems to not let you scroll to the end when you rotate from portrait to landscape and back again.

Build something that had previews not been built is unnecessary. There is a whole universe of possibilities with APIs, mashups or re-doing something with your own twist.

I'd love to see some built that is not for profit in this day and age. Not to be overly negative. Perhaps it should be build what you want, not what earns the most.

Build what had previously not been needed.

Reply on news.ycombinator.com