onfocus

Just noticed that Amazon is now allowing comments on each review. It looks like this:


For years, Amazon has tried to keep people from having conversations via reviews with careful instructions about not referencing other reviews. But that didn't stop people from conversing. You often see things like, "Reviewer x is off his rocker..." or "I don't know what some of these reviewers are thinking..." in Amazon reviews. Now people can talk to each other directly.

If people know that their reviews are "thread starters" rather than isolated posts, you could get more chatty reviews with open-ended questions designed to provoke discussion. You should also get more flame wars, more trolling, all of the standard online discussion problems. (Especially with an audience as large as Amazon's.) And how do you police comments on millions of reviews? Does each reviewer "own" the thread associated with the review? If so, shouldn't they be able to approve/edit comments on that review? Is it fair to allow comments on a review from six years ago, when the author of that review isn't expecting feedback, and likely isn't tuning into the page anymore?

On the positive side, you might get a better view of a product because discussion can bring out more detail. Should be interesting.

A few years ago I put together a little tool to help assemble RSS feeds of Amazon products called the Amazon RSS Feed-Builder. I've been using feeds generated with this tool for about three years, tracking the latest books, music, and DVDs across series and artists that I like. Because publishers often announce books to Amazon well in advance, I know about new books in the Hacks Series well before O'Reilly announces the books on their own website. Amazon also offers pre-built feeds on their Amazon Syndication page.

These old-style Amazon feeds have worked well at alerting me about new products, but they are fairly limited. I just see the title, the author, and a price in my newsreader. I decided to upgrade my Amazon feeds so each item includes a product image (if available), a product description, and product details. And I figured if I was going to go through the trouble of upgrading my feeds, why not just upgrade the Amazon RSS Feed-Builder? So here's the new thing:

Amazon Feed Generator

It's hot off the assembly line today, and I'm sure there are bugs to be worked out. (It's also powered by orange gradients.) If you want to give it a shot, feel free to try it out and post any comments/problems on this post. As an example, here are the latest books in a Polar Exploration Feed. I subscribe to this feed, and I'm notified whenever new books about polar exploration show up in Amazon's catalog.

This uses the latest version of Amazon Web Services, with a custom stylesheet and Amazon's server-side XSLT service. And I want to say thanks to Alan Taylor for his recent article subtitled, AMZN-XSLT-JSON-AJAX (AXJA?). His stylesheet is a perfect example of consuming the new, more-complex AWS responses with XSL.

ETech has been over for a week, and one presentation is still nagging at me on a regular basis. Amazon has a Web Service called Mechanical Turk (named after this Mechanical Turk), and Felipe Cabrera from Amazon spent 15 minutes or so talking about MTurk during one of the ETech morning talks.

The talk focused on the idea that artificial intelligence hasn't materialized, and there are still some tasks that are easy for humans but impossible for computers. For example, a human can look at a picture of a chair and answer the question: Is this a picture of a chair or a table? A computer would have a tough time with that.

MTurk farms out these sorts of questions to real live humans and wraps their decisions (or HITs in MTurk parlance) into a Web Services API so they can be used in computer programs. Cabrera called this process of tapping humans to make decisions for machines Intelligence Augmentation (IA) as apposed to Artificial Intelligence (AI). The talk was good, and MTurk is definitely a clever hack, but the idea has been bothering me.

I can imagine a world where my computer can organize my time in front of the screen better than I can. In fact, I bet MTurk will eventually gather data about how many HITs someone can perform at peak accuracy in a 10 hour period. Once my HIT-level is known, the computer could divide all of my work into a series of decisions. Instead of lunging about from task to task, getting distracted by blogs, following paths that end up leading nowhere, the computer could have everything planned out for me. (It could even throw in a distraction or two if that actually increased my HIT performance.) If I could be more efficient and get more accomplished by turning decisions about how I work over to my computer, I'd be foolish not to.

I guess this idea of people being managed and controlled by machines is nothing new, and it was the bread and butter of science fiction books I read as a kid. But MTurk puts this dystopia in a new, immediate context. Machines are smarter than ever, and control of human decision-making could be highly organized.

MTurk is only a few months old, and there's nothing inherently wrong with it. But I can't stop projecting the ideas behind the system ahead a few years, and that's what's bothering me. I can't even fully articulate why it's bothering me. I don't have any conclusions, or even concrete hypotheticals of MTurk gone awry—so I'm just using my blog as therapy. Obviously my computer didn't ask me to write this.

As you probably know, I put together a book called Amazon Hacks a couple years ago, and there are a couple of other books I've worked on, so I have a few reasons to watch Amazon very closely. One way that I watch Amazon is by subscribing to an Amazon RSS feed that lets me know when new books in the Hacks Series are available. On August 9th I noticed that another Hacks book I put together—Yahoo! Hacks—was available for pre-order and I posted about it. At the same time, I set up a script to take a screenshot of the Amazon detail page for Yahoo! Hacks every day at 4:10pm. I wanted to watch the page evolve, and have a record of how the page looked at various stages of development. I often think of Amazon detail pages as static, but they change quite a bit over time. Because there's no way to go back in time for individual books (unless you can go wayback on a title), there's no record of this change.

So with that in mind, here's a rough timeline of the evolution of the Yahoo! Hacks detail page I found by watching the screenshots.

August 9th [view]

• Page added
• Price is at list price of $24.95

August 16th [view]

• The Editorial Review gets better formatting
• Page count jumps from 352 to 452

August 23rd [view]

• Cover image added
• First Sales Rank ranking

August 31st [view]

• Related Listmania! and Guides added

September 6th [view]

• Amazon drops price to $16.47 (standard Amazon discount)
• Page count at 488 pages

September 10th [view]

• Editorial review has even better formatting with bullets

September 28th [view]

• Page count at 489 pages

October 19th [view]

• Buy button changed from "Pre-order" to "Add to Shopping Cart" (I posted about it)

October 24th [view]

• Product dimensions added

October 30th [view]

• "Better Together" offer added
• I added my own image of Yahoo! Hacks to the page with the "customer images" feature

November 23rd [view]

• First customer review is added (it's good! shew.)

So yeah, probably of no interest to anyone but me. But I thought I'd share what I found. By the way, Flickr Hacks is available for pre-order at Amazon now. You too can watch that page evolve.

I've enjoyed reading the conversation philosophical exhibitionism going on in the Amazon Reviewers post from last Tuesday. In trying to find latent motivations for the 1,282nd book review on Amazon, I missed a bunch of overt motivations and they've been covered well in the comments. The most obvious motivation is knowledge sharing, as in, "hey, this cajun cookbook doesn't have a recipe for gumbo!" I still don't think this adequately explains thousands of reviews for a single product, because not every review is adding a new data point about the product.

I had lunch with Lisa Ede today and we talked about weblogs, Amazon reviewers, podcasts, and more. I always leave our conversations with a lot to think about. She's studying citizen reviewers—those folks like you and me who offer opinions on products at Amazon or on our weblogs. She recently gave a talk called Online Citizen Book Reviews and the Circulation of Cultural Power and posted the text to her weblog—...my ISHR talk.

In her talk she asks a very interesting question:

What motivates someone to submit the 1,282nd review of The Poisonwood Bible to Amazon.com?

And I stumbled across a similar question on the Freakonomics authors' blog in a post called Why do people post reviews on amazon?:

Take the latest Harry Potter book, for instance. It has been out about a week. So far there are 1,385 reviews at amazon, and another 385 at bn.com. What's in it for reviewer 1,385?

I thought it would be fun to try to answer this question, because I think what motivates reviewer 1,385 or reviewer 1,282 is very similar to what motivates an anonymous blogger to post about what they had for lunch today. (I had a ham and cheese sandwich.) I came up with three potential motivations:

I think one motivation for reviews in general is building reputation, though it's not a great answer for reviewer 1,385. The chances a review in the middle of the stack will be read are low, so it's probably not a good strategy to post reviews to popular books if you want to bolster your Amazon reviewing identity. Also, it's no easy task to break into the Amazon Top Reviewer list, though it may be easier to gain fame/reputation with a body of offbeat reviews.

Another possible motivation—and you'll have to stick with me on this one—is that some might want to more closely tie their identity with a particular product. Because our identities are often tied up with the products we buy, music we listen to, books we read, it makes sense that some may just want to add a piece of themselves to an official page for a particular product. It's a bit like leaving an offering at a shrine for a particular deity. This is basically the concept of commodity fetishism, and I think Amazon benefits from this effect.

Another potential motivation occurred to me after reading about Jeremy Heigh's distinction between conversation and philosophical voyeurism on a kottke.org post in The present future of conversations. I think an act of philosophical exhibitionism can help people organize their thoughts or simply help them feel they're contributing to the common good. If I expose my thinking (as I'm doing now), and it sparks someone else's interest, or leads them down another path, I've contributed something to the world. And if I can explain to someone else why the Harry Potter book is good or bad, I make those thoughts more concrete in my own mind.

Lisa likened the 1,282nd review to the scene in Network where everyone throws open their windows and screams, I'm mad as hell and I'm not going to take it anymore. And I do think there could be a social motivation that's trying to ease a feeling of isolation.

So now that I've tried to gain whuffie, paid homage to Amazon by associating my identity more closely with it, and now that I've helped myself clarify my thoughts by exposing my internal monologue—and hopefully helped others in the process—I can end this post.

How did I miss this? Amazon is analyzing the text of books to find Statistically Improbable Phrases (SIPs) contained within. This is a funny, interesting way to get a glimpse inside the book before you buy. For example, here are the SIPs for Gladwell's Blink: rapid cognition, intuitive repulsion, sip test, adaptive unconscious. (Overheard on #ETech.) I hope someone builds this for weblogs.

Let the A9 local image hacks begin: Run up and down the street. Put in an A9 search URL, and get a panorama of the whole street. (to the left or right.)

Also, Craig let me know what the group bike thing is from my last post: a seven-person conference bike.

Amazon's search engine added pictures of storefronts to their yellow pages. For example, here's camera shops in Portland. Here's their description of how they did it. This seems like it'd be a nightmare to keep up to date, but Amazon is a huge company with lots of resources. Will they let business owners update their photos themselves? And I wonder if any of these bulk photos caught snippets of people living their lives (by accident). In any case, it's great to see the world wide web getting local like this.

Update: heh, Alan Taylor was wondering the same thing about unintentional A9 photos, and he set up a Flickr tag (a9local) to track interesting images anyone finds. (For example, what is this thing?)

oh lazyweb, why do you mock me? I'm trying to comment on a post at lazyweb, but it says:

Your comment was denied. It contains content currently banned by my blacklist.

oh yeah? Well now you're on my blacklist, lazyweb! fwiw, here's what I was trying to say—

It's not easy. I hacked my Amazon RSS tool to do UK feeds: Amazon UK RSS. A better way would be to write your own Amazon UK XSL stylesheet. People have been asking Amazon to add UK feeds for a while now. Alan Taylor put together the RSS for Amazon.com, but he has moved on, so this feature may be on the back burner for Amazon UK.

That doesn't look like comment spam to me.

MIT Technology Review has an article about Amazon Web Services up: Amazon: Giving Away the Store. I talked with the author—Wade Roush—a while back, and he mentioned some of my comments in the article. He provides a good overview of why Web Services are important for non-techies, and I really like Wade's summary of Amazon's Web Services/syndication strategy:

[Amazon] has, in essence, outsourced much of its R&D, and a growing portion of its actual sales, to an army of thousands of software developers, who apparently enjoy nothing more than finding creative new ways to give Web surfers access to Amazon merchandise—and earning a few bucks in the process. The result: a syndicate of mini-Amazons operating at very little cost to Amazon itself and capturing customers who might otherwise have gone elsewhere.

I mentioned something about Amazon's Web Services strategy being "almost frightening." I think that's becuase I've primarily worked at and with small companies. So when I see what's possible when a large organization decides to pour resources into a project, I'm impressed by what can be accomplished. It's funny to see comments I made in what I viewed as a conversation with someone end up in print, because the quotes feel out of context. (Though I don't think the quotes are distorted or misused here in any way.)

Sweet! Alf over at HubLog, hooked up Amazon.com Citations and Touchgraph: TouchGraph browser for Amazon Citations. He set it up as a bookmark, so you can launch it from any Amazon page. He also has an example set up using the book I mentioned: TouchGraph citations for Moral Animal. Since the data isn't available via the API, he must be scraping the HTML.

The coolest part: double-clicking a book reference brings that book up in the graph, and you can see which citations the books have in common.

Talk about the LazyWeb in action—thanks for the fun, Alf!

Update: Looks like the citations feature may be in flux at Amazon, so the HubLog TouchGraph browser isn't working at the moment. (It was fun while it lasted!)