Sunday, August 30, 2009

Aweekstweets August 24-31 2009 @jameskobielus


1. Recollecting the week in my mind. Pred anal key to scientific discovery. Model decay: old "laws" lose force if new data not scored.about 5 hours ago from web
JK2—Theories are cheap. They’re just predictions about what future evidence may (or may not) confirm. Only as strong as the next confirmation. Decaying into superstition with each failure—or willful refusal—to confirm against the actual universe.

2. Contemplating new gen of predictive analytics in social media. Vet "friend"'s likely future behavior online. Hedge "friendship" accordingly.about 5 hours ago from web
JK2—It comes down to reputation. If you think of it, reputation is a predictive metric. Per what I said in this blog a few years ago, reputation is “an assurance or trust level—an evaluation of the extent to which someone is worthwhile to know and associate with.” It’s “relying parties’ evaluation of our reliability, of their liabilities, and of the degree to which associating with us makes them ill at ease. Relying parties...gather assertions and data from [hither and yon] before rendering their evaluations and opening their kimonos." Before establishing an online “friendship.” I’d like, among other things, some assurance of the other party’s reciprocal friendliness. I’d like to know if they’ll respond quickly, reliably, fully, warmly, constructively with me though any or all social media channels. Nothing I hate more than gushing my guts to someone who doesn’t give as well as they take. How can I check their track record on all this without violating their privacy.

3. RT @jennifer_dubow: "my team would like to listen to your podcast on BAM/BI next week." jk--We record 8-31. It'll be posted by mid-week.3:10 PM Aug 28th from TweetDeck
JK2—She’s referring to the podcast referenced in this blogpost: Is BAM Relevant in the Age of Lean Processes?. Never worry (spoiler alert): BAM/BI’s mommy survives this episode.

4. Truly advanced analytics = complex info, models, deployments, latencies, policies, usage. Also, embedded. Hence, seamless, invisible. Simple2:53 PM Aug 28th from TweetDeck
JK2—I thought I had coined the term “transalytics” to refer to the tight integration of analytics into transactional computing applications. But then a quick Google disabused me of that notion. Apparently, my grandiose conception is a high-falutin way of saying “tape-reading.” I stand corrected.

5. RT @JesseNewhart: "Over 75% of orgs haven't upgraded to latest MSFT Office." jk--Nor should they. Core/older features sufficient for most.9:06 AM Aug 28th from TweetDeck
JK2—I’m a late adopter of most new techs, services, gadgets, apps, features. And I’m not staying up at night exploring the features of the techs I’ve worked into my life. I’m far more interested in using whatever willfully limited toolset I have to generate and share the content that continually courses through my head. Hence, the “content to be content” motto on this blog.

6. RT @madgreek65 @atmanes "Sell SOA to the business, pitch it as cloud" jk--Substitute one bubble abstraction for another. They'll eat it up.7:55 AM Aug 28th from TweetDeck
JK2—2009 feels like 1969. Saying “SOA” now is like saying “far out” then. What was cool and hip a few years ago now marks you as a dork. Remember how John Denver was saying “far out” well into the 70s? Remember when Duke was neighbors with him in Aspen in Doonesbury? “That’s a far-out creek you have there, Duke!”

7. @kitson @jowyang #AugmentedReality #socialmedia Mobile wearable devices that we all strap on? Wristwatch-like? Yuk. They chafe & constrict.7:46 AM Aug 28th from TweetDeck
JK2—Speaking of things that mark you as a dork. Any technology you wear, as opposed to simply wielding. Any interface that causes you to gesticulate weirdly in apparently empty space. Any service that isn’t woven transparently into the fabric of everyday reality.

8. Just upgraded to Office 2007. So I'm in sync with 2 years ago. But it changed my doc UI. New look, new feel. But I lost the feel. Feels odd.7:23 AM Aug 28th from TweetDeck
JK2—By definition, any new interface that doesn’t give me the option of keeping the old interface cramps my style. I’ll migrate my work- and lifestyle if and when I’m good and ready. Not when Microsoft or anybody else dictates.

9. RT @jswanhart "How do people who follows 1000s of other people filter all that stuff anyway?" jk--You don't filter. You simply glance.
JK2—Worry about information overload is so 90s. The stuff just flows and flows. Don’t worry about missing stuff. It pours directly into the ever-filling, ever-searchable “historical record.” That latter’s a misnomer, of course. The past is always presenting itself. The things you need to find will find you.

10. Eight tweets. That's the # of consecutive tweets I can page down in TweetDeck with my right index finger on mouse wheel. 8 tweets = 1 twyte.1:35 PM Aug 27th from TweetDeck
JK2—A twyte’s more than enough distance for a quick glance. I curl my twytefinger on my mouse like an inchworm on a leaf. I don’t attempt to colonize every leaf on this massive tree of info, much less the entire forest canopy, or the world’s many arboreal ecosystems.

11. Interviewed by reporter yesterday on some privacy-invading "data mining" tech. I'm a bit tired of discussing DM in "necessary evil" context9:11 AM Aug 27th from TweetDeck
JK2—I had to drill the reporter, Deborah Yao of Associated Press, for fuller context on the story, which concerned some vendor, Echometrix, that I’d never heard of. The story had something to do with Echometrix’s parental control software, installed on home PCs, feeds teenager private instant messages, tweets, blogposts, and other Web 2.0/social media contents in real-time to some online service hosted/managed by Echometrix. Apparently, Echometrix’ “Pulse” service “reveals what teenagers really think about your products and your competitors.” Echometrix collects and aggregates this info by demographics into an online data mart, and sells access to advertisers and marketers. Interesting because it mines commercially golden customer-generated content, and leverages key text analytics (sentiment analysis), but it still leaves me a bit cold. It feels like surveillance. It feels like privacy invasion. Yes, these are households that voluntarily yield this info. But what are they getting in return? And do teenagers have no say whatsoever in the decision by their parents to sell access to their most private communications (albeit aggregated, de-identified, mined for larger patterns and trends) with some unseen company and their advertiser customers? I can just imagine the privacy advocates, such as Marc Rotenberg’s Electronic Privacy Information Center (EPIC), getting wind of this. Data mining often gets presented in this privacy-invasion context in the popular press.

12. "First Look: Microsoft's SQL Azure Database CTP" ( jk--Literally: screenshots: svr adm, scripting, query opt, rsc gov..7:27 AM Aug 27th from TweetDeck
JK2—Good to see they’re well along toward production go-live of the service later this year. SQL Azure’s architecture and interfaces don’t interest me as much as Microsoft’s go-to-market business model. Are they primarily going to embed this cloud database in Microsoft Live online app services (a la what they did with the Passport authentication service several years ago)? Are they going to pitch it to their SQL Server enterprise customers, or to small and midsized customers, as a lightweight SaaS database service for specific apps? If so, which customers and which apps? I’m hearing no clear direction from Microsoft on how they’re going to compete with other vendors in the emerging cloud DBMS market.

13. RT @mbalch "Need efficient way to condense relevant tweets" jk--That's purpose of my "aweekstweets." Weekly self-digest. Tweeter's digest.6:59 AM Aug 27th from TweetDeck
JK2—It’s also a handy-dandy way for me to leverage “note to self” tweets as material for longer thoughts to be developed later. And to collect my scattered observations into more coherent aggregates. And to keep my personal blog regularly refreshed with new content, some of which may find its way into my Forrester blogs and other means of expression. Poetry needs to be just one element in my personal blog, not the whole thing.

14. Glanced at the 1000th (or millionth?) trade press article ripping MSFT a new (old?) one over "Bob." Give the snark a rest, please, folks.2:45 PM Aug 26th from TweetDeck
JK2—Ever notice the unimaginative, kneejerk, copycat, schoolyard herdthink that we all, even the most professional of us, tend to engage in? This ongoing “let’s jab another dagger in Microsoft Bob’s long-dead sucky heart” theme somehow reminds me of a point that the late biologist Stephen Jay Gould made in an essay ages ago. He discussed the essay, “The Case of the Creeping Fox Terrier Clone,” in this interview. A key excerpt: “Textbooks are so unimaginative and just copy one from the other. I discovered that in describing evolution, the horse, which is a standard example in all texts, the earliest horse called the eohippus, is always described as fox terrier in size. Now, it occurred to me that everybody who copies that line--most of them probably don't know what a fox terrier is outside of seeing Astin, those old "Thin Man" movies--who knows what a fox terrier is, yet 80 percent of all references give this knee-jerk line. Eohippus was fox terrier in size. So, I traced it, and I realized that they're just copying from other textbooks. You get these traditions of unimaginative textbook copyings. So I wrote this essay on how the unimaginative textbook is one of the main impediments to good science education.” In IT punditry, the “Microsoft Bob sucks” theme is not much more than that. Each subsequent compiler of “stupidest product” lists just aggregates the previous bunch of lists into a new list. I suspect that the older the product, the less likely the most recent stupid-product-list aggregator actually used it. Given that Microsoft Bob came out in the early 90s, its presence on new lists probably owes more to the “my big brother told me it sucked back in the day” than to any fresh analysis based on “let’s take it out of box, re-install it, and give it another look-see just to be fair.” For the record, I had Bob on a home PC and liked it. But it was limited by the fact that I couldn’t install my favorite cartoon characters for the ones that Microsoft bundled with the product. I love cartoons.

15. Noticing attempts to weave quasi-tweeting into blogs; eg. @cmooreforrester (, or my own private-blog "aweekstweets."2:43 PM Aug 26th from TweetDeck
JK2—Always looking to expand my expressive powers and not get shackled to tiresome old thought patterns. When I was young, I thought and wrote like an old man. I’m trying to make up for lost time.

16. I have entirely lost interest in the universe of Twitter add-on features/tools. I just like simplicity. Lean tweets.2:36 PM Aug 26th from TweetDeck
JK2—I hate distraction.

17. Did a great podcast with Stefan Andreasen, Kapow's CTO/co-founder, and Dana Gardner, Interarbor Solutions, on web data services.1:00 PM Aug 26th from TweetDeck
JK2—Discussed my FORR doc on Mighty Mashups: Self-Service BI For the New Economy.” I disagree with the headline in this otherwise fine article by Lauren McKay in Mashups don’t make anybody smarter. If you wish or don’t watch out, they allow you, the end user, to make an even bigger mess with mashed-up Web data than you would with spoonfed enterprise data. You still need to distill it all down to a drop or two of insight. And that’s the “heavy lifting” I allude to: the uncanny brainpower you may or may not apply to the mashup.

18. New FORR blog post: "Is BAM Relevant in the Age of Lean Processes?" ( AM Aug 26th from TweetDeck
JK2—Good. The folks at Information Management syndicated what I consider one of my best recent blogposts. Spoiler alert: The answer is “yes.”

19. "Anti-Twitter: Woofer requires 1,400-char min" ( jk--Charcount-sorted msging? E-mail evolve to route outbound by size?7:27 AM Aug 26th from TweetDeck
JK2—I pride myself on my ability to go long or short with my musings, whatever suits the message, or my mood. Novels are not a more advanced art form than short stories. They’re just longer.

20. "Social Networking Poll Shows Users More Vulnerable Than Ever" ( jk--Vulnerable to those tuned to our vulnerabilities.7:19 AM Aug 26th from TweetDeck
JK2—Users are not more vulnerable than ever. Rather, our self-appointed “protectors” are more predatory than ever.

21. Nothing could be sweeter than a clean inbox at 8:12am--i.e., a formerly full one cleaned through brute force--i.e., resolve to respond later7:16 AM Aug 26th from TweetDeck
JK2—Nothing’s more transient than a clean inbox. Don’t freak when it’s suddenly full 10 seconds later. Just keeping draining—and breathing evenly.

22. Developing predictive models is one thing. Predicting which models will be most useful is another. That's why comparative model eval so important7:13 AM Aug 26th from TweetDeck
JK2—Any darn fool can predict the future. Many predictions are plausible. Adjudicating among them is best left to the experts. Economics wouldn’t be the fine, Nobel-winning profession it is today if this weren’t true.

23. E-mail evolving into event-alerting, clickback, and object-transport bus for interpersonal comms. Social ntwks are core P2P "messaging" now.8:20 PM Aug 25th from TweetDeck
JK2—Somehow, this doesn’t feel like it’s making us more social.

24. Developed BAM maturity model. Will discuss w/ @passion4process on FORR podcast next Monday. Is BAM still relevant in age of lean processes?3:23 PM Aug 25th from TweetDeck
JK2—BAM BAM BAM! Level 1: business activity monitoring. Level 2: business action mobilization. Level 3: business adaptability management.

25. @carterlusher : Yep, re Forrester research projects (e.g. Waves), research associates are key. Mine is the ever-ready/organized Charlie Coit11:33 AM Aug 25th from TweetDeck
JK2—Even with a full-time RA, it becomes awfully difficult to manage my full research agenda, consulting, speaking engagements, etc. Being a Forrester analyst is a step or two up in the “everybody wants to speak with you” scale from any of previous, smaller analyst firms I was with. Not that I’m complaining. Being with a big-brand gives me unparalleled access, and unparalleled influence. None of which do I take for granted. That’s why I work doubly hard.

26. RT @carterlusher: Observations on Forrester's and its analysts' use of social media AM Aug 25th from TweetDeck
JK2—Forrester encourages us to use all social media channels, new and emerging, to build our respective star-analyst brands and our corporate brand. In case you haven’t noticed, I’m one of the most active tweeters among current Forrester analysts. I’m also one of the handful who have their own private blog. Which I started long before I joined Forrester—and which has obviously evolved over the years as my career has evolved—and in line with my ever-changing moods and interests. I like to play around. Call me quirky, but you only go around once. I’m the master of me.

27. "Google PubSubHubbub Looks Like RSS on Steroids" ( jk--It also looks like key pub-sub protocol for social networking.8:25 AM Aug 25th from TweetDeck
JK2—Will the ever-diversifying social-media world ever standardize on a common pub-sub protocol? Hell no. Some master aggregator service will emerge by tapping into the proprietary interfaces of all the major—and many minor—social media services. That aggregator may be Google, but probably not. Hence, PubSubHubbub could just become yet another visionary but ultimately forgotten, nowhere-implemented Google project. Remember when I referred to Google as the Xerox PARC of Web 2.0? I still stand by that.

28. RT @carterlusher: "Why large analyst firms don’t seem to mind losing superstars" ( jk--Good post. 2 words: deep bench.2:01 PM Aug 24th from TweetDeck
JK2—What’s questionable is this notion of analyst “superstars.” There are lots of smart people in this space. But none of us is so supernaturally, charismatically brilliant that they clearly stand head and shoulders above the rest and set the industry’s agenda. I judge everybody by a simple criterion: what have I learned from them that I couldn’t have figured out before long on my own? I’ve long surrounded myself with the smartest people I could find and then, in my head, quickly assimilated what I’ve learned from them. What others might regard as “superstars” I might regard as somebody who’s simply well known and respected. Down deep, the core criterion I respect is brainpower—and collegiality.

29. RT @bitterer: @hackathorn "condensing a vendor briefing into 140 chars *is* a bitch." jk--Technical term? Is a batch of bits a "bitch"?12:29 PM Aug 24th from TweetDeck
JK2—It’s not a bitch. It’s easy as can be. To the extent that the vendor doesn’t condense their message down to a simple phrase and make that their press release headline, they’re not doing their job. I laugh at how often I can do it for them—though they may not agree with the slant I give it.

30. "Mining the Web for Feelings Not Facts "( jk--Cube evolve for advanced analytics? Feeling fact & dimension tables?10:55 AM Aug 24th from TweetDeck
JK2—OK, OK, you OLAP purists—I know a “feeling table” can be modeled as just another type of “fact table.” Chances are, though, it often won’t. It’ll be modeled as an RDF triple instead.

31. Teradata Virtual Storage “Temperature Aware” [for] DW ( jk--Auto-move data to low-cost "colder" storage. Key cost-saver10:39 AM Aug 24th from TweetDeck
JK2—If a data warehouse were a real warehouse, “hot storage” would be the main loading dock where trucks pick up same-day local deliveries; “tepid storage” where goods scheduled for delivery in the next 2-3 days, within a 500-mile radius, are kept, further back from the docks; and “cold storage,” where less-in-demand inventory, not currently scheduled for delivery, are maintained. In other words, a data warehouse, like a real warehouse, is just a complex, many-sectored storage, staging, and transshipment points for diverse goods.

32. Taking a demo from IDS Scheer on their BAM and BI mashup solutions.
JK2—Also, I’m taking demos and briefings galore from the vendors in my Forrester Wave for Predictive Analytics and Data Mining Solutions. I’m not tweeting or blogging on any of those. You’ll have to wait till I publish my Wave later this year.