The TED Augmented Reality Hoax | Hacker News Reader

The TED Augmented Reality Hoax

unitzeroone.com

124 pointsllimllib50 comments

From the TED site, posted by one Matthew Trost:

I want to let everyone know that TED is working quickly to fix the missing acknowledgments.

We're speaking on the phone with Chris Hughes, who sends his sincere apologies for neglecting to mention FLARToolkit and Papervision3D on stage. When Chris showed us the software off stage at TED, we jumped at the chance to show it to our audience, and swooped him up to the stage after a very short prep time! He was also careful to mention that he based his project on the work of many others in his TED Blog interview.

This is such an astonishing development in software, and we want to make sure those who participated in it do not feel cheated. We're going to add prominent acknowledgments to the video and our blog posts so our audience is clear about the work and talent that went into the toolkits that were the basis for what Chris presented. If you have any questions or further suggestions, please contact me -- we're listening.

Sounds like a lot of noise about not much.

I'll side with Chris Hughes, against the overly pedantic, on this one. It's hard enough to say anything in 2 minutes. To say it right and without making slips is not something I'd expect of anyone other than a PR professional.

Before you say "but he clearly said that..." think of how much pressure you'd feel standing on the TED stage with just 2 minutes to make everyone go "Wow!". In those conditions, your brain will naturally be focused on something other than making sure everyone gets a mention.

Edit: for those pointing at the blog for evidence, I doubt Chris Hughes wrote that blog post himself. It was probably based on his talk and some quick chatter with the writer. So it's just an amplification of a slip.

My impression was that the "idea" he presented was not a piece of software per se, but what you can do using widely available tools. He said at least 3 times that the important part is that you can do this using a browser and adobe flash. That you also need a couple of libraries is pretty much obvious.

This being said, I'd expect that somewhere around the video to find links to the software he used. Otherwise the whole presentation doesn't really make sense.

I usually hate these tempest-in-a-teapot flame wars, but I'll actually have to disagree here. He gave a whole demo on stage claiming credit a bunch of times. I went to the FLAR toolkit page and downloaded the demo apps. This is exactly identical to the demo, just with the logos swapped. There is not one bit of original work that I can see. I'd like to believe it's an accident, but it seems like if I demoed Google with my logo slapped on it and accidentally forgot to mention that someone else wrote the search engine.

Download the toolkit and check--I'd love to be wrong, but I don't see how. And it would be unfair to pretend this is a slip a steal credit from the people who clearly spent years actually building it.

I'd like to believe it's an accident, but it seems like if I demoed Google with my logo slapped on it and accidentally forgot to mention that someone else wrote the search engine.

Perhaps I'm blind, but I didn't see his logo on there.

What I did see was a TED video being streamed over that square.

So, a more apt analogy would be: it's as if I took a laptop with a wireless connection to a part of the world that doesn't know about Google (let's call it Grahamia), and I said, "Look what we can do with this! We just type 'Grahamia' in there, and within seconds we get all these awesome bits of information about Grahamia. That's great, isn't it? Oh, my 2 minutes are over. Thanks everyone, hope this helped!"

If you have 2 minutes to get a cool piece of tech across and you decide to spend it talking about what technologies it relies on, I don't want to see your demos..

I see your point and how it's somewhat in-between but would argue that my metaphor is closer.

The TED stage has a well-known role that's clearly understood by everyone there. I would never get ask to go up on stage to demo Google. Larry and Sergey would. If I got asked to demo someone else's work, it wouldn't slip my mind. While it sounds short, 2 mins is a pretty long time.

As I mentioned, just my opinion. But I do think someone's got to stick up for the developers who wrote the sw.

Erm, no.

Here you go:

> I'd like to believe it's an accident, but it seems like if I demoed Google with the TED logo slapped on it and accidentally forgot to mention that someone else wrote the search engine.

Did they change the audio from the originally posted video? I just watched it and I didn't hear him claim to have developed anything he was showing.

Nope, I don't think they changed it. Which is why I think this is overblown.

Swombat, wtf is your agenda? Yes, the audio is changed because the new one no longer has him saying "I've written a bunch software".

If you're trying to back up a friend, you're being way to obvious. This is definitely not overblown, taking credit for other people's work should be career changing and in the downwards direction.

*Edit: the change was really subtle too, go to 0:20 on the original

http://www.youtube.com/watch?v=6CRI3X0pjaU

I didn't notice the change, my mistake.

And no, I don't have an agenda. I don't know this guy, but I know a lynch mob when I see one, and I don't like siding with the pitchforks.

I saw this too and was wondering what was new in the talk. It looks like he downloaded the source, swapped in a TED logo, and took credit. After it came out that the talk about training crows was faked too, TED is looking pretty bad IMHO.

I remember seeing the crows video. Do you have the link to where it was shown it was fake? Can't seem to find it by googling.

http://www.nytimes.com/2009/04/12/magazine/12letters-t-CORRE...

disclosure: Josh Klein is a friend of mine from grad school (ITP)

Joshua Klein's crow-training project's validity is a point of contention. The original NYTimes article was poorly researched and much was assumed/embellished through no fault of Klein. The correction the Times semi-recently published made it out to be Klein's fault, though he attempted to contact and correct the Times, to no avail. Poor journalism and self-interested/important journalists are the suck.

I was at the talk at TED and was blown away by what he said did. I spoke to some other folks at ITP and they said that he didn't actually train the crows in the way he very explicitly said he did during his TED talk.

He very clearly gave the impression that he had trained the crows following all the steps described in his talk.

I could be wrong about the details, but it certainly wasn't the NYT's fault. They reported it as he described it.

Regarding the NYTimes story: he posted a response here.

http://www.wireless.is/projects/crows/correction.php

If what he says in his response is true, then it means that some shoddy reporting and fact-checking by a NYTimes intern did an enormous amount of damage - not just to Klein, but everyone involved in his research, including his advisor and both universities.

Regarding the talk: I agree that he does give the impression that the coin box experiment has already succeeded and he's ready to move on to more practical applications, when in fact his research is still in its early stages. It's a mistake, but a long shot from what the person in the original article has done (and what the NYTimes has potentially done to Klein's career).

This seems like a lesson in too-early publicity.

Also if I ever talk to a reporter I'll take my own notes and send them to him/her and post a version on the web so it can be Googled.

Hindsight is 20/20.

I am sure Mr. Klein thought that the journalist/intern was going to practice due diligence (that would be my assumption coming from a brand like the New York Times).

I've heard of people bringing their own tape recorder when talking to a reporter. It can be that bad...

They've retracted and updated with due credit: http://blog.ted.com/2009/06/a_note_on_today.php?utm_campaign...

Just shows once again that being a good marketeer is often more important than being a good engineer.

What are the odds that Chris parlays this into a startup or a nice research position somewhere? Either way, the guy now has a TED presentation on his CV and the guys who did all the legwork have no legal recourse because I don't know of any OSI license has a clause about presenting the work as your own for non-commercial purposes.

The only way Chris will get busted is if TED decides to do so and expose themselves for their poor background checking which is a crap-shoot at best.

Code can still be copyrighted under an OSI license.

Sure but I can get up on stage at any conference and talk about copy-righted work all I like so I don't think it matters.

Copy-rights overwhelmingly are concerned with explicitly commercial uses. If I'm using copy-righted work to build up my academic reputation, I don't know if there's anything a copy-right holder can do as the link to money is too tenuous.

I think all the enforcement has to come from the academic side in cases such as these where reputation is all that's in play.

Do you feel even a little bad about celebrating this guy's bad faith?

My opinion of his actions (which I think are pretty reprehensible, for the record) don't change the reality of the matter.

The guy beat the system and that will continue to be the truth of the matter unless TED acts (which remains to be seen).

"What are the odds that Chris parlays this into a startup or a nice research position somewhere?"

Any research institution dumb enough to hire him based on work like that is probably not one worth working for. What he showed on stage was approximately the first 3 weeks of an undergrad computer vision project. Anyone who knows anything about AR knows that.

What is sad to me is that he clearly really wanted to be on TED, and if TED asked him to talk about his real work I'm sure we'd all love to watch it. But instead he spent an hour following a tutorial of some OTHER random new cool photogenic technology and TED decided that was good enough to put on their website, even though he personally added absolutely nothing to this technology.

The guy posted about it on his blog:

http://www.spazout.com/holy_smokes_im_on_ted

Oddly, he coats his statement with the same passive language bullshit that drove everyone nuts regarding the CouchDB/Porn debacle: "I realize that it should have" and "it appears that I could have".

What's so hard about saying "I was wrong not to mention the toolkit and I apologize"?

I'm starting to wonder if this is a just a problem with the writing skills of the people involved. Maybe this guy simply does not know what an apology looks like or who to go about writing an effective one. Reading this blog post, it seems like the bones are there and he stops just short of actually apologizing.

I'd love to believe that, but ultimately I can't. The most naive writer who seeks to apologize simply says "I'm sorry."

That's an earnest apology, not a naive one. Even those who feel they did something wrong, desperately want others to understand their side. I think the problems comes from people mixing in the explanation for why they did what they did with the apology.

I think it takes some sophistication to realize that if you explain yourself, there must also be an unequivocal apology separate from that explanation. See Richard Clark's "I failed you" testimony.

Even those who feel they did something wrong, desperately want others to understand their side. I think the problems comes from people mixing in the explanation for why they did what they did with the apology.

Good point. I hope it's earnest and I'm inclined to take it at face value. That being said, going out of his way to minimize all the outrage at him misrepresenting thousands of man-hours of work as his own will probably cause his apology to fail for many.

If you read his earlier blog posts about his talk at TED, he mentions it was his life dream to be/speak at TED with HIS creation. He doesn't mention anywhere that what he showed wasn't his technology - he very much makes it sound like it was his own. He tricked them.

TED should have done some research.

It looks like the talk (http://www.ted.com/talks/chris_hughes_demos_easy_augmented_r...) has been removed from TED's website. It's advertised on the front page, but goes to a 404. I think its removal is a good decision, although I wish TED had replaced it with an explanation and apology instead of merely disappearing it.

Explanation here: http://blog.ted.com/2009/06/a_note_on_today.php

Also, they have overlayed credits to the open source projects.

The note doesn't do it for me. The issue is not just Hughes's lack of proper attribution on stage, but his assumption of credit. The interview doesn't absolve him, it implicates him, when he says ARToolkit is unbelievably awesome and then, "Folks ported it from C, which is what it's written on, onto Java. And then we took the Java port and made it work in Flash." Isn't he saying that he was integral to the port from Java?

Later in the interview, he says that making AR easy was "all we were looking to do." He calls himself a researcher and says, "It's open source. Everything I do is about being very open." On the stage he says he "wrote a bunch of code". I can't come to any other conclusion than that he misrepresented his contribution throughout. But neither he nor TED has acknowledged that fact or apologized for it.

I'm not saying we need to string him up for it, but both he and TED need to stop making excuses and start making a sincere apology.

They changed it - I think it's better now.

Clearly, TED needs to vet their presenters a bit more. If a guy has no connection to the tech that he's presenting (other than he's played with a demo), then he should never have been invited.

Or we need to accept this sort of thing as a tradeoff, the price we pay for a fairly lightweight TED system. Heavy vetting has a cost, too.

Perhaps just add another commandment about giving attribution and credit (and not representing others work as your own) to the existing TED commandments:

http://www.swiss-miss.com/2009/05/the-ted-commandments.html

From a broader view, Papervision would be impossible without the thousands of hours of work other people did to create the technology it is built on, and that tech would not be possible without the work of the people who developed the tools that made those tools possible and so on and so forth back and back and back. I'm not saying it isn't a nice gesture to let people know who's hard work enabled you to combine different technologies in a new and exciting way, but lets have a little grace here becasue everything any one of us does it built on somebody else's hard work. Even if you code everything you do in assembly you are still "standing on the shoulders of giants". (Quote courtesy of Sir Isaac Newton)

I hear you, but that'd really be applicable only if the presenter actually added a material contribution or at least a novel recombination of his own. He literally just followed someone's tutorial and presented that.

I find this behavior quite appalling. Not giving credit is not only dishonest, but shows complete lack of character. The fact that Chris Hughes runs his blog under spazout.com is just another sign of his moral stance.

Edit: to clarify why I take offense with 'spaz out', I consider it a slur towards people with cerebral palsy. See also http://en.wikipedia.org/wiki/Spastic and http://en.wikipedia.org/wiki/Spasticus_(Autisticus)

I can see why you would feel that way about the phrase "spaz out", but just to be clear: I don't think that's a common view. Saying "spaz out" is not commonly considered politically incorrect, at least as far as I know; I've literally never heard that argument before.

(I'm fairly neutral on the issue at hand, btw; I didn't know of Chris Hughes or the software he used before this incident.)

I'm all for giving credit where it's due and I think he could have mentioned the libraries he was using in passing, but.. you don't have much time in a 2 minute tech demo to capture the audience, and I don't think this audience particularly cared what api's he used. He did post links to Papervision3D and FlarToolkit on his blog.

The problem is that he made it sound as though he was representing his own work, which it was not.

If he indeed was just demo-ing something he built using existing libraries and a simple tutorial he found online ... that is a far cry from being able to claim he was demo-ing a piece of software he "wrote".

If you're gonna be that pedantic, how did he create this piece of software after reading the tutorial?

Oh yeah, he wrote it.

This is like claiming credit for a LEGO kit that you built by following the intructions. You end up with something really cool, but you had no creative input. It's much more admirable to take the pieces and turn them into something new that no one has ever thought of.

Reply on news.ycombinator.com