Joi Ito's Web

Joi Ito's conversation with the living web.

My investors, my readers and a variety of other people keep trying to get me to explain what I'm interested and why I'm interested in it. Here's a first shot at this. Thanks to Steph, Kevin Marks and others on #joiito for a first pass edit. I've put it on the wiki as well so we can continue to work on this.

Context instead of content

Attention is moving from commercially produced content to dynamic or contextual content. An example of this is the shift of Japanese youth spending from CD purchasing to karaoke to cell phone messaging. CDs let you passively consume content produced by companies. Karaoke is more interactive - you are part of the content. With Cell phone messaging, the customer creates the content. From a copyright viewpoint, CDs are strongly protected. Karaoke is less protected and usually licensed in bulk, and messaging has very few copyright issues. With 20 million camera phones in Japan alone, text messaging is adding photo sharing, making conversations look more and more like content publishing. Small morsels of content, created by users and shared is called micro-content, as opposed to expensive commercially produced and protected content.

Networked consumer electronics devices will make PCs less relevant

With each new wave of computing devices, from mainframes to mini-computers to PCs to game consoles to consumer electronics devices, there is a huge increase in volume causing a dramatic decrease in cost. The users and application developers also shift to these new platforms for better performance and smaller sizes. We still have mainframes and mini-computers but they are less relevant. PCs will become less relevant as the number of consumer electronics devices with networking features increases. Eventually digital cameras, phones, TVs, PVRs and other devices will all be connected to the Internet. People will be publishing, sharing, viewing and hearing content from the Internet without having a PC. They will be as irrelevant to consumers as mainframes.

New open standards for micro-content and metadata

The third important trend is the blossoming of open standards built for creating, publishing, syndicating and viewing/hearing micro-content. Open standards have been around for a long time, but the weblog community is making them popular. These open standards are currently being tested and developed primarily for PCs, but many of the standards could be used in consumer electronics devices, allowing smaller developers to write applications and web services for consumer electronics devices. This is very similar to the way in which TCP/IP allowed the developer community to write software for communications leapfrogging the large telecommunications companies. There are many standards for consumer electronics devices, but they are complex and mired in committees, rather like CCITT's x.25 standard that TCP/IP quickly replaced in many applications.


As broadband becomes cheaper and computing power increases, everything we're learning and building around text micro-content and metadata will be useful in dealing with multimedia micro-content and metadata. Because it is more difficult to extract meaning from images and audio, metadata about this content will become vital.

So what's going to happen?

Microsoft will continue to dominate the desktop, but it will become less relevant as consumer electronics companies embrace open standards and use Internet web services and applications to make consumer electronics devices rich with content. The content will be micro-content such as photos, audio clips, video, text, location information and presence information of friends. Digital rights management and copyright will become less relevant. Organizing your network of friends and your network of trust become more important, so that you publish to the people you wish to hear you and you are able to sort information which is relevant to you. These trust networks will require privacy and security as well as methods of managing and using the networks for a variety of applications.

As web services and metadata create a more and more decentralized and semantic web, searching will become more decentralized and contextual and less about html page scraping and one dimensional page rank.

In the future, you should always be able to see the status of your friends (if they choose to let you), create any kind of content you wish to share or communicate and publish it easily from any device. You should be able to find and view/hear any content you have access to, using your network of trust, location, keywords and timing to search for the information. The boundaries between email and web publishing will become blurred and you will be having conversations with the web.

Key Technologies:

  • Creating and managing identities while protecting privacy
  • Creating and managing networks of friends and trust
  • Searching metadata and creating context for metadata
  • Design and interface for publishing and viewing micro-content
  • Syndication standards and technologies
  • Network infrastructure to enable location and mobility
  • Technologies to move and share micro-content, especially as it grows larger
  • Web services that interact with micro-content and the physical world such as photo printing, purchasing of real world products, connecting people, etc.

The cutting edge:

Audio blogging (Audblog), mobile picture blogging with location information (Tokyo Tidbits), personal information and information about your friends in web pages (FOAF), machine readable copyright notices allowing micro-content aggregation and sharing (Creative Commons), Amazon book information and affiliate information embedded in blogging tools ( TypePad ), convergence of email and micro-content syndication (Newsgator), searching for micro-content based on context (Technorati)


I just blogged this but I thought I'd share it here too:
It's a good summary, but it misses a key point. By emphasising the difference between commercially produced 'content' and user-created 'micro-content' he is ignoring the enormous area inbetween.
The current content publishing model is only efficient for large-runs of sales - sell under a few thousand books, a few hundred thousand CDs, or a few million cinema seats, and you won't be welcome in commercial publishing.
This gap is gradually being bridged by innovative companies, such as Cafepress and Customflix, but both of these are still creating physical goods.

I think there is a huge opportunity here to be the eBay of digital media, and I think mediAgora is the way to go about it.

Of course the trick is getting this all to happen in a world of lots of broke programmers. Right now - unless you have a viable business model and your earning money already - you can't get anybody to invest in any of these key technologies.

Where are the investors who will help make all this happen?

Marc, I'm not so sure it's a bad thing to say that entrepreneurial programmers need to bootstrap themselves to some extent. Why is it a bad thing to have a "viable business model," with evidence of potential success, before you seek funding?

Obviously, Joi is one of the investors who will help make all this happen. The Movable Type example is a good one--the Trotts had a viable business model, and evidence of money-making potential.

Seems to me that investing in programmers with ideas, sans business models or demonstrated revenue potential, is part of what got us into the dot-com disaster to begin with.

(All of the above should be taken with a grain of salt, since I'm neither a programmer nor a successful investor. ;)

Nothing wrong with viable business models - it's just that no where in Joi's thoughts does he bring that up. These thoughts are all about the technologies, the ideas and the raw, potential building blocks.

Those building blocks imply innovation, but in fact - the marketplace doesn't pay for innovation and what's getting funded are in fact - pretty traditional models: blogging tools, Wiki consulting companies, mating services. So if innovation and cool, key technologies are what Joi is interested in - does that mean he's investing in the technologies or the companies that implement and deploy those technologies?

In fact from where I'm sitting each of those key technologies is a part of a unified, digital lifestyle kind of product - that no one company or product can deliver. Only Apple and Microsoft can fund, develop and deliver products with that scope. And each of the key technologies Joi mentions can't stand alone as viable models.

What has to happen is that tools have to hook up with social networks, cell phones have to blend with blogging tools, micro-content has to be syndicated, meta-data and web services have to all be presented as an integrated solution. Probably the most important thing about ALL of what Joi says is that digital identity has to have CONTEXT? Who's gonna pay for JUST their Digital Identity?

So in other words - all these 'technologies' are pieces of the puzzle and they need to be united with the business models, management, leadership, compunction and zeal that makes an entreprenuer who she/he is and the startup successful.

And THAT combo of the elements is what needs to get funded. Without matching components, management, timing and funding - we'll never have the kind of results we all hope to see happen. THAT's what should get funded!

I certainly agree that viable models are important, but if you believe as much in what Joi seems to believe in - then perhaps a little faith and belief in the people who have to DO all the work is important too. You know - the people who build the companies.

Kevin: Good thoughts. I'm going to figure out how to integrate them into my thinking. Would like to chat about this with you some time.

Marc: Agree. Writing general thoughts is the first step and not enough. Coming up with how to put these thoughts together into a business model is harder. Then finding a team to execute this is even harder. I think the execution bit is 90% of the work.


"Where are the investors who will help make all this happen?"

In a sense, I think it's good. Often when an investor has given money for a project, it has killed the project. I have this idea, but I may be naive, that a project needs maturity to evolve.

A plant takes time to grow. You certainly need to take care of it, add water, and remove old bits, etc. But too much money and requirements on a project and the project disappears.

I would like business angels, not giving money to project but management and organization and let the things grow slowly. Go forward to the next step when you have made a balance.

The problem with this model is... business angels want often fast profit... which IMHO, is not compatible with a good business model.

Of course, chaotic solutions are the ones that mirror the internet's structure, and are going to find a more organic and fertile soil than any single solution. What's missing, IMHO, are models for human cognition that can engage these spaces. We're still largely living in a Newtonian world in our heads... not even near the 20th C. And what about some model of economic granularity to engage the commerce of microdata et al.?

I've been having this conversation for 2 years now with the operators and ISPs here in Taiwan/China. It's like talking Martian to goldfish. Joi's concept is about the way things will emerge. It's not a business model, or even a market yet, but a vision of what is building up. Not the "me too" mentality that runs most organizations over 25 people. There is a business model here, but I don't believe it is based on any existing value chain, because value chains once defined are useless because the real business opportunity is in the non-chain or the up and coming chain which unfortunately most people with power and money cannot see, and depend on do-ers and movers to tell them where it's actually happening. Do you ask a record company exec where the latest music is coming from? No way. You ask some kids in the scene. They know because they breathe it. It's the same for this kind of tech stuff. You have to play with it to know what's really going on. And eventually people will bank it, but by that time we'll be doing something else already.

Good post. I'm not convinced that meta data will become more important; rather, some equivalent of Google will find out how to decipher the "meaning" and origin of various pieces of flotsam content and will categorize it accordingly.

Users won't label or tag content, simply because labelling is time-consuming and doesn't offer them any benefits or advantages. Users just want to communicate instantly.

-- mike

Mike: The tools we produce have to create much of the meta-data automagically. Consumers will see LOTS of benefits - once the semantic web weaves it's quilt.

Corbett: It's building the new value chain and market that Joi is talking about. We all need to work together to do that.

Jason N: YES - it's not just ONE killer app, it's actually the process of integration, aggregation and appropriate levels of customization. Everything we need has been invented, it's just time to get it all to work together.

Karl: Believe me I know what you mean by investors meddling and killing companies. Because multimedia didn't take off the way they thoguht it would (by '90) the VCs took over my company in '91. But you still need energy to fuel the fires. You can't keep asking people to work on spec. Investment is needed to implement Joi's roadmap.

There isn't just one business model here. There are potentially a whole bunch of them. Major new applications like these develop from the bottom up, they aren't imposed from the top down.

Neither Apple nor Microsoft can develop these products. The reason is that these products are individually too small and do not encourage the hegemony of a single company. Neither Microsoft nor Apple is flexible enough to respond to the many different needs this marketplace will potentially serve. Don't wait for the big guys, do it yourself.


Good post. I agree that personal contextual personal content will become king and that "personal networks" will zooooom. The amount of personal content that will be published and shared will quickly outpace professional content over the next couple of years.

A couple points that you should be thinking about (and something that should influence how you invest your money):

Broadband Internet, other than relatively narrow and expensive DSL connections to the home, is stagnant. This will push more of the intelligence to the periphery of the network. Unlike other places in the world, there isn't any government effort in the wings to rectify this. We are going to be stuck until high bandwidth wireless networks bail us out at the end of the decade.

In the meantime storage and local wireless LAN technology are moving ahead down the price/performance curve much faster than any other technology in computing. Additionally, these technologies can be bought and implemented by individuals without waiting for centralized (RBOCs) decision making.

This means centralized services based on the Internet/Web will likely be much less attractive over the medium term. Instead, software and devices that leverage these rapidly moving technologies to allow people to create personal "sharing" networks will zoom. Think real P2P apps that soak up all the bandwidth a paltry all you can eat DSL connection provides. Think wireless person to person local networks. Think decentralized.

Individually constructed metadata falls prey to the problems so eloquently described by Cory in metacrap.
However collaboratively constructed emergent metadata is very useful. This is the Hayekian point about price as an information system, or the way Musicbrainz gathers song information, or imdb movie information, or wikipedia gathers articles.
You need an agreed canonical ID structure, and the ability to update and improve, and a heuristic for deciding what constitutes an improvement.

John is correct about the importance of the edge but the best articulation I have seen is from Ray Ozzie.

As far as zones of trust Weblogs are like front stage while groove is backstage about sharing secrets. Both are communication and collaboration network management tools that leverage the value contained inside social influence networks.

When can I have groove on my cell phone? Cell phones also leverage the value of the network. The value is more from the possible connections that cell phones enable. Likewise, in messenger, sometimes I flip through my messenger buddy list or groove contact just to see the alerticons and think about who I can message even if I don't intend to actually chat them up any time soon.

I agree with Kevin on the commercial opportunities that stand between individual publications and traditional commercial distributions. Our self-syndication Media Guide targets these independents.

Regarding media playback and format shifts (text to audio) a related thread is going on in the Open Source Media Forum on There the challenge is to build an Open system to document the 2004 US elections resulting in standardization of some of the concepts discussed here. While the Elections system needs to be purely Open, many of the proposed methods sit comfortably with commercial endeavors.

Will PCs really become as irrelevant as mainframes? Seems just as likely they will stick around and become a "hub" for users' other devices, possibly as a storage device for all of the photos/text/audio/video content that people are creating. Tools from companies like Apple and Microsoft seem likely to play a big role here - Apple leading the way. Of course innovative new tools will be created (ex: Audblog, Typepad) but ultimately isn't there still a HUGE incentive for desktop sw giants to commoditize the personal content creation tools market as others were?

hi joi, if pc's become irrelevant how will i maintain the ability to access so much porno? also i would like to take this opportunity to predict the increasing irrelevance of capital letters.

Joi is shining a giant flashlight which I think many aren;t focusing on. He is predicting, albeit conservatively, that the usage and usability of digital communciation tools is going to explode in as many directions as we can imagine. The problem for the engineer or tool developer is not how to get funding, but which tool to build first. The needed and desired tools in the next five years will certainly include enterprise-scale apps, but the vast majority will be garage-and-all-nighter-scale using costless open source tools, quick design and peer-review. Of course this latter group will create few insta-riches or web-famous, but they will provide for hundreds and hundreds of individuals to cover their expenses and a buffer cushion to make their next product.

Similarly, for the content producers, you will have dozens and dozens of more tools to use to sell and share your wares. The real promise of global audience and distribution is already flowering all around us. Local bands are selling songs around the world. Popular custom clothing and books and art is sold thru websites that cost less than $25/mo to run. Quality products will be found by desiring consumers. Quality products will be paid for.

From my perspective if you want to sit around and wait for investors or critics and "important" bloggers to validate what you know, then you're gonna miss this round of opportunity too.

"There is no try only do." Yoda

Re: the PC thing... I don't think they'll go away, just like mainframes haven't gone away. They're workhorses for many processes. I just think that when the volumes of smaller devices increases SO much and the power increases with it, it will make more sense to write services and software for these devices instead of focusing on writing stuff for the PC. It's also A LOT EASIER to get paid for stuff on consumer electronics devices. Young kids in Japan happily pay $100/month for data services on the phone. No way will they pay that for Internet content/access. People happily click "subscribe" to cable TV or other services on their TV that they would never buy on the Internet. So much more reason to write applications and services for CE's (Consumer Electronics devices). More customers, more low cost computing power (including cheap cameras, DSP's, etc.), and people who are willing to pay.

What this will probably do is cause many software companies to start developing for CE's instead of PC's and venture money and talent will also shift.

Something similar happened a bit in the game industry when a huge chunk of programing talent moved to the console game market and sucked energy out of the PC game market.

Well I guess I would agree that the PC as a development platform is becoming more and more stagnant as time goes on and the energy will move elsewhere - but one caveat is will there be ANY type of standardization provided by the platform vendors on cell phones, pdas, etc. The nice thing about Windows is its one (sorta) set of APIs to write to, whereas things get murky on these newer platforms. And don't tell me WAP is the answer.

Of course Joi is speaking from a place (japan) which is probably 2 years ahead of us here in the States, in terms of mobile technology. So my skepticism is to some extent a function of being in the U.S.

But ultimately I do see major barriers due to the desire of each mobile carrier to have their own proprietary standard - should I write to the Sprint network, or Verizon? Japan is lucky in that iMode is a de facto standard. This lack of standards is a real problem for small app developers who want to use something other than SMS though. Its unclear how basement developers make money today off of mobile standards.

Hi Joi, I liked Amar Bhide's point in "The Origin and Evolution of New Businesses" that promising start-ups don't come with viable business plans because they are in areas of "high uncertainty" as opposed to "high risk". Examples are the origins of Microsoft and HP. You simply proceed opportunistically without trying to measure the risk. Our lab is working on "business ecosystem" and "fractal team building" models for leveraging such high uncertainty We're currently writing this approach into a proposal for the Campaign for Better Transit in Chicago, citizen assessment of government performance. Also, we're developing a "knowledge radio" for monitoring wisdom of investments. Peace.

How about the Open Source developement method? I've just started a wiki for discussing Social Software done the Open Source way:

Do you think there will be more common acceptance worldwide of SMS text micro-novels i.e. Qian Fuchang, Joshi, Phil Marso et al?

16 TrackBacks

Listed below are links to blogs that reference this entry: Thoughts on micro-content, metadata and trends.

TrackBack URL for this entry:

joi ito on what's interesting him from the anti-mega outboard brain
July 22, 2003 5:01 PM

Joi Ito's Web: Thoughts on micro-content, metadata and trends A nice roundup of why things are changing... Read More

Well, that's a grandiose title. Sorry. I am not going to try to provide a complete courseware critique here. I'm just thinking about one thing that bothers me about the courseware we use at RIT (and which is true of most courseware systems)--it's close... Read More

Joi Ito's Web: Thoughts on micro-content, metadata and trends Thought provoking article by Joi which illustrates a lot of key points about the technology trends within the blogging community right now. Perhaps even more interesting are the comments on ... Read More

And also Joi's Thoughts on micro-content, metadata and trends... Read More

Joi Ito's Web: Thoughts on micro-content, metadata and trends Interesante opini〓de Joichi Ito sobre micro-contenidos y metadata. Merece la Read More

Keli l〓ir eftir traustum Linux net〓〓 til a〓tengja fars〓a vi〓til framb〓r, e〓 a〓to〓fr〓Landss〓anum og Vodafone vi〓a〓reka SMS bloggi〓 Gummi J〓er a〓sko〓 hva〓 t疚n SMS-〓t 〓up〓 g疸t Landss〓ans 〓lir Hann s... Read More

This is were things get a little bit technical.... So... it appears that Landssiminn & Vodafone have a sms 2 email gateway up and running. It's a service that's been on my todo list for some time, basically trivial to... Read More

Two interesting blog entries about content for mobile phones: For mobile video think context, not content Thoughts on micro-content, metadata and trends When you think about the current state of the economy it doesn't sound like a good business to... Read More

I found this post by Joi Ito to be really interesting, especially in light of what we're currently doing here Read More

Joi Ito: Thoughts on micro-content. "So what's going to happen? Microsoft will continue to dominate the desktop, but it will become less relevant as consumer electronics companies embrace open standards and use Internet web services and applications... Read More

Extraits d'un post exposant les tendances en mati鑽e de contenu 駘ectronique et sa g駭駻alisation, par Joi Ito : "Les PC Read More

Tech Predictions for 2004 from Tech-Weblog by Christoph C. Cemper
January 2, 2004 7:02 AM

Happy New Year 2004 I saw my weblog top 40000 unique visitors (wow! more that 50% direct visitors consuming RSS feeds or visiting via Bookmark are even more amazing I think) which continues the trend I already found and which... Read More

joi ito谈技术趋势joi ito在他的文章 Thoughts on micro-content, metadata and trends 中谈到了对于微内容、元数据和技术趋势的一些想法,可以参考。他谈到的技术趋势如下:1)关联文本将取代(单向)内... Read More

Why am I going thru all this hassle of exporting content from blogger and haloscan? The simpliest explanation is... Read More

When Content Takes The Back Seat from WorldChanging: Another World Is Here
June 10, 2005 10:04 AM

You may have already read that Google is now the world's largest media company, in terms of stock capitalization. (Which is not the same... Read More

Microlearning emerges when new media keep breaking the world into. Work, knowledge, communications … all is falling into small digital fragments, loosely joined and permanently rearranging to form a multitude of new patterns, tasks and threads. We ha... Read More