Update: The BS News story was scrubbed, no explanation given. Story, explanation, and pictures are here.
ONE OF THE few things worse than being woken up by an alarm clock at 7am after getting to a hotel well past midnight is being woken up before 7am by a phone call from a very angry Wolfgang Gruener. That was the first time the dirty underside of plagiarism in the IT journalism world directly made my life interesting, but far from the last.
The frustrating part about plagiarism is that it tends to be subtle and hard to prove, there is usually more than enough plausible deniability to go around. In this case, Wolfgang, editor of TGDaily, was as much an unknowing victim of it as The Inquirer and I were. This and several other incidents have one thing in common, Theo Valich, proprietor of BS News.
The first such incident that I had ever heard of was said to have involved Theo and Scott Wasson of The Tech Report. It happened years ago, and no one that I talked to can remember the details, but some are alluded to in the chat log below. The only reason I recall it is that there was a second such incident in the middle of August, 2007.
Intel, like just about every other company out there, briefs journalists before a product is launched so they can ask questions, write things up sanely, and review parts where applicable. Depending on the part, it can go from an email or phone call, to a meeting at a company office to play with parts on site.
The audiences are also different, sites/magazines that care about a GPU probably don’t care about technical server minutia, and storage oriented sites are unlikely to care about gaming mobos. The smallest bunch of them all tends to be the server group, especially for deep technical dives. At these briefs there are about 10-15 journalists, the group never changes much, and we are a very odd lot.
I attended one of those briefings in the summer of 2007, the topic was on the revised Intel quad core server parts, chipsets, interconnects, and other bits that are tied into the platform. These briefs usually consist of two days of dense technical talk, lively Q&As and in general, more numbers and graphs than most people can digest in that period of time.
If you are not at these briefs, you tend to get a slide deck that has all the relevant information, but not the same level detail on a few topics. The slides are more than enough to get the needed information about the product, and if there is anything missing, Intel is very good about answering any questions.
Scott Wasson wrote up his usual excellent article on the subject here, and as is often the case, looked at an angle that few if any others did, idle power. Due to an editorial mix-up, someone at The Inq wrote a brief version first, and that went up here. The next day, Theo Valich wrote a follow up piece on the new CPUs, and it is here.
On the surface, all looks well. Scott and I were at the Intel brief, and we got our information there, Theo probably got the slides from Intel later on and filled in gaps in the initial Inquirer piece. Sounds good, right? The problem is that the piece that Scott picked up on, idle power, was fairly obscure, and was nowhere in the slides given out to those not at the briefings. In fact, it wasn’t in any slides that we got, if you didn’t take notes, you didn’t have that information. Oops.
Theo’s original article did not credit Scott Wasson or The Tech Report at all. Scott was rightly angry at this, he did a lot of work, had a unique angle, and felt ripped off. If my memory serves, when challenged, Theo claimed that he got the information from Intel slides, but could not produce them when asked to do so.
Paul Hales, editor at The Inquirer at that time, added the link at the bottom of the story retroactively, it was the right thing to do. I initially defended Theo, he was a friend, and when your friends are accused of bad behavior, you naturally stand up for them. I found the situation odd, but up until the point where Theo could not produce any document or link that showed the information, I didn’t think anything was amiss. Reporters legitimately get the same information from multiple sources all the time, but they are always able to back it up, publicly or privately.
Step forward in time about six months to March 20, 2008. Theo Valich was no longer at The Inquirer, he had moved on to TG Daily. I got off a plane in Austin, TX after midnight, the flight was badly delayed. The clock was into the single digits before I got to the hotel room, and we were leaving for briefings on the upcoming IALP, now Atom, CPUs at 8am. A 7am wake up call would be about enough time to shower, eat breakfast, and stumble to the front desk for a ride to Intel.
My cell phone woke me up at 6:45am, and seeing as how the calls I get at that time tend to be emergencies about servers going down or related network problems, I answered. I had met Wolfgang Gruener a few times before, but I can’t recall ever talking to him on the phone until that point.
I was extremely tired, trying to get my eyes to focus, and here was Wolfgang apoplectic with rage, asking me why I was attacking him, and why I had put “that story” up. I was barely coherent, Wolfgang also seemed barely coherent, and I had no clue what he was talking about. The last story I put up was one about AMD layoffs, and while it may have bothered AMD PR, it was unlikely to have thrown Wolfgang into a frothing rage.
As I slowly woke up, and Wolfgang calmed down, I found out that he was upset not about my story, but this one. Reading it, I can understand why he was upset, to get called out in public like that has got to be unpleasant, accusations of plagiarism on this level can ruin a career.
When I logged onto the net and followed links, I read this story on TG Daily, and I was quite amazed at the similarities to my AMD layoffs one published a hair over three hours before it. Please note that at the time, the story had Wolfgang’s name on it, it was written and published by Theo Valich, but due to a glitch or a wrong button pressed, it ended up with Wolfgang’s name on the masthead.
Now the situation was coming into focus, even if my vision hadn’t quite caught up. I wrote the story, Theo published the same information hours later, Paul Hales noticed it and thought the two eerily similar. Too similar in fact, so Paul fired off a broadside that landed in Wolfgang’s lap. For some reason, all sides wanted to talk to me about it, quite a joy in the pre-dawn hours after a night of almost no sleep.
I told all parties that while I thought the information was exclusive to me, there was a possibility that Theo could also have gotten it from the same source. I said that I would circle back, talk to everyone involved, and try to put the pieces together. With tempers calmed and promises made, I took a shower and headed down to breakfast.
Ironically, the first person I saw sitting down to eat was none other than Scott Wasson. We had a long chat about plagiarism, history, and Theo. Needless to say, I got an earful, and soon was realizing the mistake I had made defending Theo over the Xeon idle power story above. Live and learn.
I started digging for the layoff information that very day, and continued for the next month. I called everyone I knew, asking who knew what, and when. Questions about what level of executive had access to that info and when it was spread internally were posed and usually answered. The short story was that very few, startlingly few, knew that it was coming, and it was not widely spread at all.
My source was adamant that he/she/it/they (hereafter he for brevity) did not talk to Theo, and did not tell anyone but me. Given our long standing relationship, I trusted this person to be honest with me, and there was no reason to doubt their denial.
One more bit to toss into the ring, a few days later, Fudo posted more or less the same story with different numbers. The reason for the differences is that the official word inside had been that there were two rounds of layoffs, about 5% each. One was more or less immediate, the other set for fall, 2008. If you had a source that was officially in the loop, you would have been told the 10% number that Fudo published.
My source had only heard about the immediate layoffs, hence the 5% number, which means that Theo’s source had to have also heard only the partial story. In any case, I took a month questioning everyone I could get in front of, and every single avenue that I could dream up for this information to get to Theo was explored, none of which panned out.
To me, it looked conclusively like Theo had stolen that story from my article, and that Paul was correct in calling him out. Given the severity of the charges, and the fact that Theo was a good friend, I thought I would give him a chance to tell me how he got the information directly to clear things up. It was the right thing to do.
On the evening of April 24, I had the following conversation with Theo over Skype. It is unedited, and all extraneous bits are still in place, as are spelling errors, embarrassing statements, and general chat log stupidities. The conversation continued on after this, but on a different topic. It is long, but I felt that it was better to include it without edits than to post bits that can be taken out of context.
[Begin chat log]
[4/24/2008 9:05:07 PM] Charlie Demerjian: You here? Got time to chat?
[4/24/2008 9:05:15 PM] Theo@TGDaily: yep
[4/24/2008 9:05:19 PM] Theo@TGDaily: not for long, but I am here…
[4/24/2008 9:05:27 PM] Theo@TGDaily: whazzap?
[4/24/2008 9:05:32 PM] Charlie Demerjian: Are you going to be around for 15 min at least?
[4/24/2008 9:05:37 PM] Theo@TGDaily: yes, of course
[4/24/2008 9:05:45 PM] Theo@TGDaily: I’ll be up for as long as we talk, no worries
[4/24/2008 9:06:06 PM] Charlie Demerjian: OK, this is a very serious thing that I didn’t want to have to do, but I promised several people I would, you being one of them
[4/24/2008 9:06:21 PM] Charlie Demerjian: Short story – the AMD layoffs thing from a few weeks ago
[4/24/2008 9:06:25 PM] Theo@TGDaily: ok, sounds very serious
[4/24/2008 9:06:51 PM] Charlie Demerjian: At the time, we both came up with the same story
[4/24/2008 9:06:57 PM] Theo@TGDaily: yes, true
[4/24/2008 9:07:23 PM] Charlie Demerjian: I told everyone that we could have talked to the same source for it, and I said that I would talk to the source when I was able
[4/24/2008 9:07:43 PM] Theo@TGDaily: was the source any of our friends we might know
[4/24/2008 9:07:52 PM] Charlie Demerjian: I did very recently, and he told me directly that he/she/it did not talk to you about it
[4/24/2008 9:08:01 PM] Theo@TGDaily: because, I learned of them from a person not workign for AMD, but still in the loop
[4/24/2008 9:08:21 PM] Theo@TGDaily: he/she is the same person that is the base for a story that will go online
[4/24/2008 9:08:25 PM] Theo@TGDaily: tomorrow
[4/24/2008 9:08:29 PM] Theo@TGDaily: and concernes Puma
[4/24/2008 9:08:42 PM] Charlie Demerjian: So, I called around and asked AMDers about it, and they either didn’t know it or had the 10% number
[4/24/2008 9:09:01 PM] Theo@TGDaily: I had “around 800 people have to go number”
[4/24/2008 9:09:08 PM] Theo@TGDaily: (mostly trimming down Texas)
[4/24/2008 9:09:24 PM] Theo@TGDaily: timeline was the only odd thing, because I had no idea you had the story…
[4/24/2008 9:09:32 PM] Theo@TGDaily: I don’t visit INQ anymore
[4/24/2008 9:09:49 PM] Charlie Demerjian: OK, normally I would NEVER ask you this, but this is the second time that I have heard accusations about this, the first being Scott Wasson about a year ago.
[4/24/2008 9:10:00 PM] Charlie Demerjian: Between you and I, who told you about the layoffs?
[4/24/2008 9:10:17 PM] Theo@TGDaily: You know him as well as I know him
[4/24/2008 9:10:34 PM] Theo@TGDaily: Scott’s accusations came after MSI gave him the same story as me
[4/24/2008 9:10:42 PM] Theo@TGDaily: Biostar too…
[4/24/2008 9:10:49 PM] Theo@TGDaily: (AMD 690G BIOS story)
[4/24/2008 9:11:18 PM] Theo@TGDaily: When it comes to layoffs, we are both speaking to the same person about things
[4/24/2008 9:11:28 PM] Charlie Demerjian: That is what I was afraid you would say. I know a lot of people like you do. The one who knows and had that specific number flatly denied talking to you
[4/24/2008 9:11:42 PM] Theo@TGDaily: hm…
[4/24/2008 9:11:59 PM] Theo@TGDaily: Silicon Valley-based
[4/24/2008 9:12:01 PM] Theo@TGDaily: that’s all I will say.
[4/24/2008 9:12:15 PM] Charlie Demerjian: Are you sure?
[4/24/2008 9:12:33 PM] Theo@TGDaily: I moved to Silicon Valley for a reason
[4/24/2008 9:12:45 PM] Charlie Demerjian: The reason I say this is because I told Wolfgang and Hales I would talk to my source, and if he said no, I would tell them
[4/24/2008 9:12:54 PM] Theo@TGDaily: Ok.
[4/24/2008 9:12:56 PM] Charlie Demerjian: He/she/it said no.
[4/24/2008 9:13:06 PM] Theo@TGDaily: You’re free to tell them that he/she/E.T. said no
[4/24/2008 9:13:18 PM] Charlie Demerjian: I really didn’t want to , I was hoping you would have a different name.
[4/24/2008 9:13:53 PM] Charlie Demerjian: The accusations in our industry are VERY serious, and for teh first time, I defended you to the wall, the second, I spent 3 weeks verifying things
[4/24/2008 9:14:14 PM] Theo@TGDaily: I know that allegations are a serious matter
[4/24/2008 9:14:55 PM] Theo@TGDaily: Frankly, whatever you, Fudo, Nova – say can be considered wrong or right
[4/24/2008 9:15:13 PM] Charlie Demerjian: OK, well, that is what I needed to know. I appreciate the time, adn sorry I had to ask
[4/24/2008 9:15:21 PM] Theo@TGDaily: All I know is I spent $450 on gas driving around and finding stories out
[4/24/2008 9:15:33 PM] Theo@TGDaily: Second story that will hit the streets soon considers Puma
[4/24/2008 9:15:34 PM] Charlie Demerjian: 10 gallons?
[4/24/2008 9:15:38 PM] Theo@TGDaily: (LOL)
[4/24/2008 9:16:01 PM] Theo@TGDaily: well, bear in mind that 1Gallon is $8 here
[4/24/2008 9:16:14 PM] Charlie Demerjian: Good point :)
[4/24/2008 9:16:33 PM] Theo@TGDaily: so, paying 3.41-79 (grew by 50 cents from my arrival to my departure) is not such a hard thing
[4/24/2008 9:16:42 PM] Theo@TGDaily: But all in all, I met with more than 50 people
[4/24/2008 9:16:58 PM] Theo@TGDaily: touchign Intel, AMD, HP, Nvidia and their respective companies
[4/24/2008 9:17:27 PM] Theo@TGDaily: and these stories are the ones I am been writing about
[4/24/2008 9:17:42 PM] Theo@TGDaily: RV770, AMD+Mubadala tie-up, etc…
[4/24/2008 9:17:48 PM] Theo@TGDaily: where I need to give credit, I give it out
[4/24/2008 9:17:55 PM] Theo@TGDaily: (you were credited in AMD-MAD analysis)
[4/24/2008 9:18:00 PM] Theo@TGDaily: (for instance)
[4/24/2008 9:18:07 PM] Charlie Demerjian: I do realize that, and thanks
[4/24/2008 9:18:16 PM] Theo@TGDaily: Credit Fudo when he does something…
[4/24/2008 9:18:22 PM] Theo@TGDaily: credit nova, credit scot, kyle
[4/24/2008 9:18:30 PM] Theo@TGDaily: (I even linked to HardOCP once :))
[4/24/2008 9:18:48 PM] Charlie Demerjian: I told both editors that I would check and tell them the answer to their questions. I checked, and I will tell them what I found without comment.
[4/24/2008 9:18:56 PM] Theo@TGDaily: ok, no problems
[4/24/2008 9:18:59 PM] Charlie Demerjian: That is about it.
[4/24/2008 9:19:13 PM] Theo@TGDaily: you can say you spoke with me now (4:19AM my time) and that’s that
[4/24/2008 9:19:32 PM] Theo@TGDaily: what happens now is that I am checking all of my stories with the INQ and Fudzilla before I run them
[4/24/2008 9:19:44 PM] Charlie Demerjian: I understand, and thanks for the time
[4/24/2008 9:19:47 PM] Theo@TGDaily: if there are no stories of that topic in first five lines (15 news), I am running it
[4/24/2008 9:19:52 PM] Theo@TGDaily: But Charlie, honestly…
[4/24/2008 9:20:03 PM] Theo@TGDaily: what bloody point would I earn if I would rip somebody’s story?
[4/24/2008 9:20:10 PM] Theo@TGDaily: This industry is what goes around, comes around
[4/24/2008 9:21:37 PM] Charlie Demerjian: I agree, but I said to Wolfgang that I would check with my source and report back. It took me weeks to get in touch with him/her/it/they, and I did, and I will write that back. No commentary, no nothing, just the simple answer to that question that they are waiting for
[4/24/2008 9:21:45 PM] Theo@TGDaily: that’s all I have to say… and we all have issues with sites like VR-Zone, Gameguru, TechAmok etc ripping us.
[4/24/2008 9:21:52 PM] Theo@TGDaily: ok
[4/24/2008 9:22:11 PM] Charlie Demerjian: I am _NOT_ doing this to attack you, I am doing this to defend you
[4/24/2008 9:22:23 PM] Theo@TGDaily: no worries, I already spoke with Wolfgang, and he and hales are communicating when it comes to subject of touchy stories
[4/24/2008 9:22:36 PM] Theo@TGDaily: Of course, Charlie, I know… everything is fine by me, no hard feelings
[4/24/2008 9:22:46 PM] Charlie Demerjian: This is the second time I have defended you, and I put my rep on the line both times for you. I don’t want to _EVER_ have to do it again. :)
[4/24/2008 9:22:57 PM] Theo@TGDaily: ok, point taken ;)
[4/24/2008 9:23:32 PM] Charlie Demerjian: Anyway, I will check one more detail and probably mail to the guys tomorrow sometime and CC you of course.
[4/24/2008 9:23:38 PM] Theo@TGDaily: ok
[End chat log]
While I was annoyed that Theo did not name anyone, we have always shared important information in the past, the detail I wanted to check was asking the source one last time about who was told what, and when. It took me a few days to reach him, and I relayed the gravity of the situation, spelling out that if he was trying to protect Theo, the best way to do it is to tell me if they talked. He again denied it.
The next day, I called Wolfgang to tell him what I had found. I was stunned when he told me that Theo had already approached him and named the source preemptively, something he refused to do for me. That said, I was really relieved to hear that this was over.
Wolfgang said that the name Theo had given him was a person in AMD PR, but would not tell me who it was. No problem there, I completely understand that part, but I was very surprised because my checking indicated PR was not in the loop when I wrote my story. Additionally, they surely would have known the real 10% number if they were told about the layoffs.
I asked if Wolfgang had verified the information with the PR person in question, and he said he had not. I asked Wolfgang to do so, and tell me what he found so I could put this matter to rest. I started calling him every few days for the next two weeks asking if he had verified the information, pestering like a child in the back seat of a car whining, “are we there yet?” again and again.
After about two weeks of this, Wolfgang told me that he had gotten a hold of the PR person in question and the information that Theo gave did not pan out. The PR person was not the source, nor was the very different source hinted at in the IM logs above. Those were the first avenues I checked out when I started looking for Theo’s potential sources.
So that is where it ended. I told Hales and Wolfgang that I would track this down, and I did, very thoroughly. It was quite clear who was telling the truth, and that is where I left it. All involved parties are adults and understood business more than well enough to deal with the issue on their own. That said, it is where my friendship with Theo Valich ended.
That brings us back to the current day, and two other stories. Read this one on The Inq by Sylvie Barak about the AMD/ATI Evergreen code names. Then read this one on BS News, Theo Valich’s site. Very similar indeed, but Theo’s is a day later than Sylvie’s. So far, so par for the course, the information could have come from a multitude of sources. Once something is officially released, it tends to travel fast.
From what I understand Sylvie published those code names shortly after visiting ATI in Markham, and scored a world exclusive. Great story, and good find. We published some of it earlier here, but nowhere near the full list. Although I personally knew most of the names, I was asked not to write them, so I did not.
The problem? Sylvie made a mistake, a very minor one, and one that I have done several times. She either misheard the code names, or transcribed them wrong, something that is easy to to when you are getting an information blizzard at a briefing. This is not meant to be a criticism of that very minor mistake.
The problem in question is that Hemlock, listed by Sylvie here, the “….low level offerings named after the poisionous[sic] shrubbery, Hemlock.”, is not a low end part, is the highest end part. Hemlock is the code for the dual Cypress board, likely to be called 5870X2. The low end parts are Cedar and Redwood, mid-range is Juniper, and the high end single chip is Cypress. Hemlock is above that, a dual Cypress/X2 board, not the low end.
Before I jumped to conclusions, I talked to AMD and asked the people that Sylvie had talked to if they had given her the information as I knew it, or as she had written it. We all agreed that her story was a mild transcription error, nothing more. I have done many of these as well over the years, everyone has. I also confirmed that there is no disinformation campaign floating around, no leaked slides with the wrong info, just the correct information. More to the point, Sylvie was the only person to officially get the names so far.
So, did Theo Valich get the right information and make the exact same mistake in this story, or did he plagiarize? I can’t think of any reasonable way that he could have gotten the same wrong information, but I am more than willing to listen to the explanation. Given the history between us, what I am not willing to do is do it behind closed doors this time.
Plagiarism is a huge problem on the web, and erstwhile professionals should not tolerate it, much less do it themselves. As writers, we need to take a stand on this, not doing so affects us all negatively, and weakens the profession as a whole. This is my stand, and from now on, when I see this behavior, I will call it out publicly. There is no reason not to give credit where it is due, and to falsely claim something is yours when it is not.
There may be a reasonable explanation for the Evergreen story, and I look forward to hearing it. Theo, your turn.S|A
Latest posts by Charlie Demerjian (see all)
- Globalfoundries 7nm process isn’t even close to the name - Sep 26, 2016
- ARM upgrades realtime offerings to v8-R and adds Cortex-R52 - Sep 21, 2016
- Everspin and Globalfoundries team up for embedded ST-MRAM - Sep 15, 2016
- Intel’s Xpoint is pretty much broken - Sep 12, 2016
- ARM adds 2048-bit vectors to v8A with SVE - Sep 7, 2016