Disqus - Latest Comments for alexiskold

Re: Monetize The Audience, Not The Content

kidmercury — Sat, 25 Jul 2009 14:15:50 -0000

i am surprised you are pro paid content. dont you think semantic technologies like your company adaptive blue benefit from having more free content to synthesize? or at least that is the way i see it, i tend to think of companies like yours as being a part of the profitable business models that will emerge dependent upon free content.

Re: Monetize The Audience, Not The Content

fredwilson — Sat, 25 Jul 2009 13:15:20 -0000

I agree but it has to done in a way that is consistent with the way the internet works.

Another thing we need to work on is payments. If the NYT asks me to pay via amazon or paypal (ideally both), I'd appreciate that very much. I don't want to whip out my credit card everytime I pay for something on the net

Re: Monetize The Audience, Not The Content

alexiskold — Sat, 25 Jul 2009 12:25:21 -0000

I am glad we are at least willing to re-visit paid content, I think this is the simplest model, that has worked for a century.

We all pay for quality stuff and quality content is still scarce, it's the subpar content that is abundand.

I hope we converge on the world where major newspapers are syndicating best relevant bloggers and have a model where they are able to pay them and at the same time, they pay their own journalists.

Re: Monetize The Audience, Not The Content

alexiskold — Sat, 25 Jul 2009 12:20:45 -0000

Jeff,

my take is that the only way local papers can compete is by generating highly relevant, high quality local content. This content is as scarce as any other content (I am not talking about crap, but quality content).

Why wouldn't you want to pay for that?

I think the model holds, what does not hold is scale, but really the scale was never the same for local newspapers.

I think Fred is right, and at the very least, this model needs to be explored.

Re: louisgray.com: There Are Two Phones In this World: iPhone and Not iPhone

alexiskold — Tue, 02 Dec 2008 11:05:11 -0000

So true. iPhone is the best gadget I've ever used. I love it in an unhealthy way. And I am not a gadget person.

Re: The need for an open library of semantic terms

ceonyc — Wed, 26 Nov 2008 12:16:51 -0000

"getting publishers to annotate is not easy, because of the lack of
direct benefit to them"

...unless people like you start directing traffic to them, or it
improves their SEO.

Re: The need for an open library of semantic terms

alexiskold — Wed, 26 Nov 2008 11:42:16 -0000

Hi Charlie,

This is good conversation on important topic. A couple of things:

- There is a question of what exactly is being annotated. Several formats address the issue in a different way. AB Meta is focused only pages that are about things. RDFa and other semantic standards offer a way to embed semantic meta data into pages. So do microformats but in a more limiting way using CSS classes. Andraz from Zemanta and a few other folks have been working on semantic tagging.

All of the above refers to publisher annotating the pages.

At AdaptiveBlue we believe that getting publishers to annotate is not easy, because of the lack of direct benefit to them. This is why we developed technology that recognizes stuff in pages in a top-down (algorithmic) way.

Regardless, whether the page is annotated by the publisher or content is recognized the next question is what do tools do with this information?

StockTweets wants to link to their site, but another stock service might want to link elsewhere. The point is that given the reconized context there is a set of actions that makes sense. Further, different actions are interesting to different users. For example, you might want to go to Yahoo! Finance for Stocks and I might want to go to Google Finance.

Tools like Glue and now Ubiquity from Firefox address this problem, by offering a set of contextual links based on the content that the user is interacting with.

To sum up, we need to have a common way of annotating different types information in pages or extracting existing infiormation out of the pages and then providing a set of contextual behaviors based on that content.

At AdaptiveBlue we have a framework, which is extensible and flexible and supports the following:
- Defining types of concepts (for example, you could define a Job Posting)
- Ability to recognize these concepts around the web
- Ability to define a set of contextual actions around these concepts.

We would be very happy to kick off the conversation about opening up all of these and incorporating things like Ubiquity.

Alex

Re: A Time Without Exits

alexiskold — Sat, 22 Nov 2008 19:09:20 -0000

Yeah, its def not simple. It seems to me that a lot of the infrastructure that was solid in the past is now in question. I am hoping that we are evolved and smart enough to quickly adapt new laws and regulations that makes sense and actually in modern times.

Re: A much happier half-marathon

alexiskold — Sat, 22 Nov 2008 17:43:20 -0000

Thats awesome, you rock! :)

Re: Contextual Browsing: Music

alexiskold — Thu, 15 May 2008 08:51:03 -0000

Fraser,

I do not remember when was the last time that I was so excited about discovering new music. I buy a lot of it via iTunes at least once in a couple of weeks and I pick stuff that I really like from the stream of NEW.

But the 4 songs that I bought from the last album I listen to over and over again. There are magical bits about them that I can't pin point, but I love them.

Re: A Conversation About Context

jonknight — Fri, 09 May 2008 21:09:25 -0000

which is strange, when you consider exactly what that does to (not for) their future business. The bits are like air: if you lock them up in your own little room they become stale and in the end, worthless.
Conversely, and also like air, if you allow them to circulate freely they are always refreshing and self-replacing.
Bits are to be used, not owned. N'est-ce pas?

Re: A Conversation About Context

alexiskold — Fri, 09 May 2008 20:24:42 -0000

Good thoughts, as always, Fraser :)

I have two distinct points about this

1) Its not really about what is the right way. Increasingly it is more about what makes people happy. People are going to consume information in heterogenous, odd ways that make them happy.

2) The idea of pulling different conversations into 1 place is a solid one. In programming there is a concept of Model-View pattern, where a model is an underlying data set and the view is one way for looking at the data. We have evolved to the point where distributed conversation on the web is the model, and each of us is looking for an individual view - a lens, or perspective through which to view it.

We need aggregators that aggregate and let us seamlessly emit thoughts back that end up at the right places across the web. The problem? Its hard to do technically and there is little incentive because fundamentally every business wants to be the owner of the bits.

Re: Introducing AB Meta - Simple Annotation for Pages About Things

egh — Sun, 04 May 2008 13:13:37 -0000

> We said that either dc.creator or book.author is supported. Frankly, people find book.author much more simple and understandable compare to dc.creator when talking about books.

Even if this is granted, going down the path of having a separate name, but the same semantic relationship, for differing types will a nightmare. Are you going to next have script.playwright? And movie-script.screenwriter?

Concepts that we use in everyday life sounds good, but I don't think this will work out in the end.

As for eRDF and your relation to the semantic web, you really need to fix those examples and clarify your URIs. Your documents are not valid eRDF.

Beyond that, eRDF seems to be pretty dead. Why not RDFa or GRDDL?

Re: Introducing AB Meta - Simple Annotation for Pages About Things

alexiskold — Sat, 03 May 2008 16:16:02 -0000

This is not a proprietary fork. Simply, this is light-weight format that has
a chance to be used by people.

Re: Introducing AB Meta - Simple Annotation for Pages About Things

alexiskold — Sat, 03 May 2008 16:13:43 -0000

Hi,

We said that either dc.creator or book.author is supported. Frankly, people find book.author much more simple and understandable compare to dc.creator when talking about books. The language matters. We already have specific semantics and instead of re-inventing it with commonly shared obscure terms, we should use concepts that we use in our everyday life. this will make publishers more amicable to publish meta data.

Re: Introducing AB Meta - Simple Annotation for Pages About Things

alexiskold — Sat, 03 May 2008 16:11:45 -0000

Hi there,

While meta tags are indeed 1-1 we have plans to expand this into microformats, and then you will be able to express more than one thing per page.

Re: Introducing AB Meta - Simple Annotation for Pages About Things

alexiskold — Sat, 03 May 2008 16:10:54 -0000

Hi Andy,

Can you please point out what specifically in Dublin Core exist to support basic everyday things? Also, the book.author or wine.winery format is meant to exactly extend things in the right way.

Re: The Week That Was

alexiskold — Fri, 02 May 2008 22:19:04 -0000

Wow, Ryan's widget rocks!!

Re: Thank You

alexiskold — Thu, 01 May 2008 20:33:55 -0000

Just want to add my 2cents on this - the party rocked in so many ways! There was this great vibe and smooth tune to the whole long evening. What a wonderful way to make us feel welcome as a company. We have so much good will and friends around us! Thank you, this helps us succeed.

Re: Party Pictures

alexiskold — Thu, 01 May 2008 20:32:07 -0000

Andy, wow! These are amazing. Thank you for taking them. My favorite view shot is the bottom one.

Re: AB Meta

Pete — Wed, 30 Apr 2008 20:48:21 -0000

Alex,

I don't know about microfromats they are not a de jure standard, just an easy way of embedding info into html.

But RDFa is a standard. It utilises xmlns just like rdf. There is bound to be a xmlns that deals with books out there(I don't know what it is and am not going to spend the time to find out though). And if there isn't you can create one, thats the beauty of rdf.

The important thing is to follow the rdf subject -> predicate -> object format. This can be embedded in html using xhtml+rdfa which is now a w3c standard.

Pete

http://localhero.biz/

Re: What We Can Learn From Beckett Baseball Card Monthly

adamposey — Fri, 25 Apr 2008 16:16:07 -0000

I have to disagree, Alex. Actually, when it comes to smartlinks I don't think either is valuable enough. The absolute standing shows popularity, but as any good hard look at the iTunes top 10 list will tell you, popularity and quality do not necessarily go hand in hand.

However, I think that showing 'momentum' in a way that provokes interaction with the smartlink would let me find the quality in it.

Re: What We Can Learn From Beckett Baseball Card Monthly

alexiskold — Fri, 25 Apr 2008 15:37:14 -0000

A momentum is the signal to buy or sell, but it is not an indicator of overall quality. Absolute standing is better at that.

Re: What We Can Learn From Beckett Baseball Card Monthly

Fraser — Fri, 25 Apr 2008 14:34:50 -0000

"There is only 1 metric that people can focus on, not many" --- rings very true.

In the examples sited above it seems that context is grounded by the absolute value and the interesting piece of information is the simplified presentation of the derivative.

I'm a new Netflix user but the absolutiness (if I may) of the top 100 is of no interest. Look at the top 10. Not interesting. What is interesting are the big jumps near the bottom of the list.

Re: What We Can Learn From Beckett Baseball Card Monthly

alexiskold — Fri, 25 Apr 2008 14:26:32 -0000

Here is a simple one - make a decision for me, just tell me - watch this movie. And if you are right all the time, then I love ya. But you know how recommendation engines pan out...