Electrons and Atoms | Hacker News Reader

Electrons and Atoms

blog.samaltman.com

73 pointsdude_abides28 comments

It might be useful to look at the history of this bits vs atoms idea. During internet 1.0 era, about 20 years back, the prevailing idea was the everything was becoming bits and the winners will be those who will manage the bits instead of being stuck on the atoms side of things. If I remember correctly, the MIT guys (negroponte) and Marc andressen (ironically, who is quoted by Sam) and others were huge into this idea. So, looks like we are coming to the idea that it is bits and atoms. I would say take it further and it will be atoms only. The bits side of get absorbed into the atoms side of things and will no longer be a differentiator.

I'm sure the first pizza place that installed a phone or fax machine was killing it with customers. But that didn't make them a communications company. Soon, every pizza place got a phone and a fax machine and then it back to who made the best pizza. I believe we are in the installing 'phone/fax' machine phase of smartphones. After 10 years it will be not a differentiating factor and the people who control the atoms side will win.

EDIT: There are companies that are digital by nature (information, music, publishing etc.) so obviously this absorption of bits into atoms idea extends to companies that have a large atoms side to them.

>> After 10 years it will be not a differentiating factor and the people who control the atoms side will win.

If you look at amazon and ebay, they are doing great. And they started mostly at the information side(but with time acquired better fitting atoms). On the other hand many physical retailers are suffering.

It is interesting that you mention amazon. The Internet 1.0 version would be amazon.com consisting of a bunch of programmers who would hook into the digital infrastructure of manufacturers, logistics providers and customers. Then by applying some fancy math they would be able to provide amazing value to customers who would then shop exclusively at amazon.com. At that point amazon would charge a ‘vig’ for every transaction on the Internet and become very successful in the process.

Needless to say, that hasn’t panned out. Amazon has invested huge amounts of capital in its physical infrastructure. Everyone thought they would kill best buy. Best buy is making a resurgence (look at its latest report) and anecdotally I see best buy having better prices than amazon. Internet company price match is now common among many big box retailers (Best Buy, Fry’s, etc.). Amazon is going to invest a huge amount $12 Billion in capital in the coming months. Co-incidence?

So, long story short, it appears that many of the latest internet companies are beginning to realize that you need a deep physical presence to make things work (see Bonobos). You can’t just build a fancy website, run some machine learning algorithms, do some A/B testing and conquer the world ☺. Look at it this way, Uber has raised $350 Million. Will the digital component (website, algos etc.) be more than 10% of that? I doubt it.

People who control the atoms side will win, provided that they have the latest atom gadgets. Or will the pace of technological innovation slow?

From my list of top startups:

Heres the ones that strongly supports it:

  1. Square: Easy payments for the real world.   
  2. AirBNB: Lodging on the real world.    
  3. Pinterest: What I like of the real world.    
  4. Twitter: Whats going on right now in the real world.  
  5. Palantir: Intelligence of the real world.    
  6. Uber: Transportation in the real world.

Startups against it:

  1. Dropbox: Storage for my digital life.  
  2. Evernote: Notes for my digital life.    
  3. Box:  Enterprise collaboration.    
  4. Splunk: Visualizations of digital things.    
  5. Cloudera: Hadoop for Enterprises.

Not sure if it's very strong - seems like they are equally distributed.

I'd put Pinterest and Twitter in the same category as Dropbox and Evernote. Evernote: What I note of the real world. Dropbox: Bring places from physical device A to physical device B.

But if you think it through, the whole list is kind of arbitrary too. Microsoft started as electrons-only company and was hugely successful.

I'd describe it less in terms of atoms and electrons, but in value delivered. And obviously, people derive value from both atoms and electrons. But they derive value differently from different services.

So, hugely successful companies help people to get value from their service and achieve something like money, fame, emotions (music, movies, art). More value for users = more success for the company.

If you accept Evernote in the atom category, you pretty much have to accept anything. Evernote notes are just bytes, and the way you interact with the app doesn't depend on whether the notes are about the real world, imagined worlds, future plans, or lunatic ravings. It's just bytes.

Bringing "physical devices" into the description of Dropbox doesn't actually imply anything; what else are you supposed to store files on? What else is any app supposed to run on? "Look at your friends' pictures on a physical device". Welp, I guess instagram is an atom company too. You've extended a useful, incomplete heuristic into a perfect, useless tautology.

I think the more useful definition of an "atoms company" is one that directly tells you which atoms to move and where to move them. This handily covers Uber, Amazon, Airbnb and very faintly Facebook, as expected, while mostly excluding Dropbox, Evernote, Google (search), etc.

You've extended a useful, incomplete heuristic into a perfect, useless tautology.

No, I just pointed out that the list of my parent poster is somewhat arbitrary. I'd put Twitter, Facebook, Pinterest, Dropbox, Evernote, Instagram all in the second category, because they exist only in digital life.

My atom life is complete without Facebook for instance. Whether I use Facebook or not makes not much difference. But booking a place via Airbnb has a direct (and only) result in the atom world: I now have a social acceptable reason to stay at a stranger's place. And furthermore: There's no other way to book that stranger's place except for maybe walking up to random strangers and asking them if they offer me a couch for 50 bucks.

I think the more useful definition of an "atoms company" is one that directly tells you which atoms to move and where to move them.

I mostly agree. But then it also depends if you take action in the atom world based on information you gathered from the electron world or not. And if you go down this route, the whole heuristic is completely nonsense, because people take action based on information gathered from electrons all the time.

I am not sure about Dropbox being in the second group. I can take a picture on my phone, return home its on my laptop and I can print it out and hand it to someone.

Likewise I can create a document on my laptop and save it in Dropbox. Later I can show it to someone in a restaurant on my cell phone.

Both are examples of bridging the real with the virtual world. Stuff I do on a weekly basis in interacting with my 97 year old non-digital Dad.

Heroku and GitHub were the first two that crossed my mind.

And they're "against it" in my interpretation of the idea.

Having said that, they do facilitate adding stuff to the first list (i.e. they are utilized by startups building services that do something in the real world).

EDIT: Added clarification of my cryptic reference to the "first list".

I think that github, heroku and others are simply positioned as providing services which the start ups who are "eating the world" are utilizing.

Yeah, that's what I meant by the last sentence :)

> From my list of top startups:

You know, a bunch of those aren't startups anymore, they're just businesses. Once you've found your strategy and are profitable, you're not a startup anymore, you're a business.

Twitter, Square, and Dropbox aren't startups anymore, they've been around for a while and are simply ordinary businesses.

You are missing the point. Call them however you want, startups or businesses, makes no difference. The point is I didn't just seek out to "proof" Sam was wrong - that would miss the point too. Instead, I followed the same steps he did.

He made a list of "breakout companies" and noticed most of them were startups that bridged the digital world with the real world. In my case the distribution was uniformly.

Perhaps do the same experiment yourself? Debating what is or what isn't a startup is not as valuable.

I wasn't disagreeing with your point; I don't even care about your point because it's as misplaced as the OP's. You're both self selecting companies you think are cool and then noting the pattern of real world vs not as if such a self selected list is in any way representative and not purely selection bias. It's silly, you see what you want to see based on a self selected list of "cool companies". It doesn't teach me anything of value.

> Debating what is or what isn't a startup is not as valuable.

Says who? It's valuable to me, I'm tired of seeing everyone call every damn company a startup cause it isn't so. A startup is a company looking for a viable business model who's still burning cash looking. It's not a profitable business who's been around for years. Making that distinction, while a side tangent, is worth making.

You don't get to tell me which part of a conversation I find interesting enough to engage on. That's my decision.

I'm tired of seeing everyone call every damn company a business cause it isn't so. A startup is a company designed to grow fast. It's can be profitable business who's been around for years. Making that distinction, while a side tangent, is worth making.

Very mature. /s

what about altmans facebook example? where do google, yahoo, heroku, github, wordpress, engine yard, groupon, admob, tumblr, twitter, foursquare...fit?

I specifically carved out enterprise software and developer tools--I think this mostly applies to consumer Internet.

As mentioned, Google is a good counterexample. Groupon and Foursquare very much touch the real world. I'd argue Yahoo made a mistake of not being enough in touch with the real world. Tumblr and Twitter are also reasonable counterexamples, although both were started before I think this trend became very strong.

Groupon and FourSquare both clearly fit. As does Yahoo with its publishing. The rest all at least arguably fit in developer tools (except admob and tumblr). Twitter facilitates live reporting of events.

I find this post to be a little bit uninformative. Sam is making the argument that the most successful companies are those that bridge the physical and digital world and increase a human's efficiency to complete a task. The startups that do best are those that focus on this.

This seems like a bit of a trivial statement in my mind. Is this not what most startups are trying to do? Solving an inefficiency in the world.

Before the last five years, it has been difficult for PG's definition of startups to focus on the physical world, because of our discomfort with putting parts of our physical world online. Most successful startups up until now have helped build up an infrastructure that has allowed new startups to emerge that tackle physical problems.

I.E. Uber required Apple to develop the iPhone and a lot of startups rely on Heroku and Parse as a service.

I really enjoy Sam's writing, and I was excited for a post dealing with the advent of hardware startups with the success of 3D-printing. I didn't find this post to be up to par with some of his other posts.

I'd also observe that the initial companies that took on the purely digital side of things (e.g., Second Life), might have just been too early in a world not yet as digital as it is today. That is, people didn't have enough of a digital life that a service aimed exclusively at it would be of sufficient value. That could change as more of life becomes fully digital. So it's possible that the online/physical thesis is "fighting the last war," so to speak and future companies might not want to pursue that.

I'm still trying to figure out something that cannot be linked to the real world

Fictional entertainment consumption?

jdileo: Looks like your account's been dead (improperly, seems to me) for at least 986 days.

jdileo: I think you have to make a new account

Wow, how do I fix?

Real estate is "real world" in big way. Most American's have majority of net worth in home/equity. As an industry, it remains terribly inefficient with incompetent Realtors costing users (consumers) real dollars. Billions of dollars of real estate each year is transacted using Realtors who do less than 1 sale per month, have limited market knowledge, no credential deeming them worthy negotiators, etc.

What is the "uber of real estate"?

Zillow & Trulia are just content sites that afford advertising opportunities to Realtors. Redfin will never achieve mass because sophisticated buyers & sellers will not---nor should they---trust these industry newbies with such responsibility. (Note: Redfin almost exclusively hires the wet behind the ears)

Interested in HN thoughts on this & it's correlation to Sam's post.

Reply on news.ycombinator.com