Friday, November 01, 2013

Aweekstweets October 26 to November 1 2013: the week old Lou turned that universal blue

"Cognitive Mining" ( ) JK--Provocative concept. Cognitive computing as interactive learning accelerator 4 human experts

"Sizing Up Data 4 CRM" ( ) JK--360-degree view dimensions: data contxt, data connections, data resolution, data density

"Open data: Unlocking innovation and performance with liquid information" ( ) JK--Links to free McKinsey report

Many will bail rather than fail. But the best know you must stay in to win.

"5 World Problems Too Big 4 #BigData" ( ) JK--Nonsense. Not too big 4 human ingenuity. Smart people use any tool @ hand

I like the fact that "wonk" is "know" backwards. If you're a true wonk, you know your topic backwards and forwards.

"Fail faster"? I prefer "know better learn faster" ( ) from Fairfax County native Thao Nguyen. #BigDataBytes

"The Data Economy Manifesto" ( ) JK--Interesting POV from @jeffreyjkelly. See my recent q-h: 

Looking forward to what @jeffreyfkelly & @stu from Wikibon have to report from #strataconf & #hadoop world on today's #bigdatabytes.

Watch @jeffreyjkelly & @stu discuss Big Data Week, preview #ibmiod. 10 min to live videochat. Watch  #BigDataBytes
Retweeted by jameskobielus

"#Hadoop & #BigData 'Stratafied'" ( ) JK--Beyond MapReduce. Into in-mem & machine learning.

"U.S. Joins Global Effort On Open Data" ( ) JK--UK-based Open Data Institute. Internatl open data net. Priv & pub orgs

“2014: Year of #BigData Applications"? Watch today's #IBM #BigDataBytes videochat, 2pm (EDT), w/ @jeffreykelly @stu .

Getting every detail locked down 4 #ibmiod next wk in LV. "Locked down" a relative thing, such complex event. Adjustments on fly inevitable

New #IBM jk blog: "Cognitive Computing: Relevant at all Speeds, Scales and Scopes of Thought" ( )

RT @kexpplaylist #kexp GMF by John Grant from Live on KEXP from Iceland Airwaves 2013 JK--Such a great song. Why's this Yankee in Reykjavik?


Google's barge. Why does that sound ominous?

Big Science? Thriving on big, open, reference data ( ) Friday #IBM quick-hit

"2001: A Space Odyssey"? Good, but it could have used a few bug-eyed monsters.

Chris Christie should make Kris Kristofferson his running mate in 2016. Just because.

RT @BradBennett: "Do You Have a Bias for Action? ( ) JK--Euphemism for "impatient"?

RT @kexpplaylist #kexp Blood Red And Goin' Down by Tanya Tucker from What's Your Mama's Name JK--This lady sang with tough sexy authenticity

Don Slack plays the best oldschool C&W on #kexp "Swingin' Doors." One awesome well-sung well-played full-of-feeling song after another.

RT @kexpplaylist #kexp The Blizzard by Jim Reeves from Th Essential Jim Reeves JK--1960. Beautifully captures th feel of trudging to comfort

RT @kexpplaylist #kexp Riders in the Sky by Vaughn Monroe from Stampede! JK--1949. Kinda makes U want 2 ride the range & learn how to lasso

RT @kexpplaylist #kexp Titoli by Ennio Morricone from The Ennio Morricone Anthology JK--1964. From "Fistful of Dollars."

RT @kexpplaylist #kexp Phantom Limb by The Shins from Wincing the Night Away JK--2007. My very favorite song from the Shins' very best LP.

Just drafted quick-hits for all 5 days next week. Yes, I'll be doing it while doing #ibmiod. And writing the following week's while there.

Intelligence also traditionally requires a pulse. But Turing wasn't suggesting plugging in an electrocardiogram.

RT @kexpplaylist #kexp Some Kinda Love by The Velvet Underground from their 3rd LP JK--Lou goes deep into a quiet, intense blues-folk vibe.

A5: #P4SPchat U dont hav 2 believ in "singularity" or any such hoohah 2 recog that cogn sys will some day advise/inform every human decision

RT @SmarterPlanet: RT @stevehamm31 A5: Computer visionary Licklider predicted man-machine symbiosis in 1960  #P4SPchat
 Retweeted by jameskobielus

RT @stevehamm31 A5:Re collaboration. Interesting findings at #MIT Center for Collective Intelligence.  #P4SPchat
 Retweeted by jameskobielus

A5: Some #MIT studies involve crowdsourcing, with computer agents in the crowd.  #P4SPchat
 Retweeted by jameskobielus

A5: #P4SPchat Humans develop algos/modls powering cogn systems; evaluate/adjust behavior of cogn sys; feed data 2 cogn sys; & consume output

A4: #P4SPchat @randal_olson Disagree. That's "Turing test not proving consciousness." Intelligence/intuition not same thing as consciousness

A4: #P4SPchat What's "intuition" but fuzzily predictive hunches that proved true? What if machines mill these hunches w/out fail? Intuition?

A4: #P4SPchat Here's "Turing test" for cognitive system creativity: can humans be fooled into thinking a machine creation from human artist?

 Retweeted by jameskobielus

A4: #P4SPchat @bobtomato28 Humans are sensors and actuators for cognitive machines, plus are cognitive "agents" in their own right. Partners

An interesting view that we become sophisticated "sensors" for cognitive machines #P4SPchat
 Retweeted by jameskobielus

#P4SPChat A4: Interesting MIT white paper about possibility of computer creativity and intuition 
 Retweeted by jameskobielus

RT @stevehamm31: A4: Humans and machines will create together. Example: #IBM work with ICE cooking school on recipes #P4SPchat
 Retweeted by jameskobielus

A4: #P4SPchat Living brain's intuition & creativity are emergent phenom from mind/body's overall sensemaking. Machine/human I/F can amplify

A3: #P4SPchat U might think of sys v. human as central vs. periph nervous sys in distrib cogn computing arch. We are eyes, ears, touch, act

A3: #P4SPchat Brain connected to entire body via peripheral nervous sys. Cognition leverages extraordinarily efficient sensemaking apparatus

A3: #P4SPchat Computing systems leverage massive parallelism & agile analytics to emulate hyperconnected fabric of central nervous system.

Wow! RT @DharmendraModha A3: Human brain consumes 20W – same computation on today’s computers would require a nuclear power plant! #P4SPchat
 Retweeted by jameskobielus

A3: #P4SPchat Computing sys emulates but evolution endowed living brains w/super-powerful/agile processing in compact, energy efficient form

A2: #P4SPchat Cognitive computing is the convergence of #HPC and #bigdata in service of human knowledge augmentation.

A2: Imagine a new era of cognitive computing where neurosynaptic chips become essential to the world. #P4SPchat
 Retweeted by jameskobielus

A2: Imagine tiny object/event specific recognition appliances for fire, traffic.  #P4SPchat
 Retweeted by jameskobielus

A2: Imagine sensing the environment with sight, sound, and smell powered by solar.  #P4SPchat
 Retweeted by jameskobielus

@SmarterPlanet A2: acting as advisor to anyone who cannot possibly absorb all knowledge, insight to make optimal decision #P4SPchat
 Retweeted by jameskobielus

A2: Imagine a rolling search-rescue robot - dynamically switchable cameras and inertial “vestibular” system  #P4SPchat
 Retweeted by jameskobielus

A2: Some #cognitivecomputing app concepts that @DharmendraModha + team came up with:  #P4SPchat
 Retweeted by jameskobielus

A2: #P4SPchat Cognitive computing enables humans to query a massive, dynamic knowledge corpus with extraordinary agilty. #Watson

A2: #P4SPchat Cognitive computing radically acceler8s "slow cognition" per Daniel Kahneman; my recent LinkedIn post: 

A2: #P4SPchat Cognitive computing's major apps include natural language processing, sentiment analysis, pattern recognition, & signal intel

A1: #P4SPchat Cognitive computing enables machines and humans to accel each other's learning, per my recent article 

A1: #P4SPchat Cognitive computing takes AI into the era of #bigdata, decision automation, and service-oriented virtualization.

A1: #P4SPchat Cognitive computing can learn from new data rather than rely 100% on apriori knowledge of data models, schema & semantics

A1: #P4SPchat Cognitive computing can auto-distill sense/signals from new, unfamiliar data sources without supervision or training.

A1: #P4SPchat What's new & different about cognitive systems is they drive machine learning on unstructured #bigdata into operational apps

Ready to start the #P4SPchat .

Cool guitar give-away contest at #ibmiod. Details in this blog:  . Rock it!

RT @SmarterPlanet Can cognitive systems help addrss #BigData challenge? Pls join discussion #P4SPchat 10-31 12noon ET 

At 12noon (EDT) I participate in #IBM #P4SPchat on "Cognitive Computing." Please join in and tweet smart.

RT @DCorrigan More than 100 sessions & activities focused on Info Integ & Gov of #bigdata at #ibmiod next week--so many choices! #bigdatagov

"#BigData University Surpasses 100,000 members!" ( ) JK--#IBM program is go-to site 4 data scientist education+training

RT @nprnews ESnowden Has Job At Russian Website, Lawyer Sez  JK--I'm sure he'll safeguard their intellectual property

"Gaming Industry Turns 2 #BigData 2 Improve Gaming Experience" ( ) JK--Stream computg, graph models, sentiment analysis

"#Hadoop 2.0 Makes #BigData Even More Accessible" ( ) JK--Good graphic showing YARN overview.

"Network Analysis of Relationship Status on Facebook" ( ) JK--Graph models can infer who you're sleeping with.

"Data Mining of Informational Stream in Social Networks" ( ) JK--Interesting slideshow of Twitter semantic graphs.

"Intel's Futurist: We'll Soon Be Living In Computers ( ) JK--Really? None o our customers has taken up residence in one

Wouldn't it be ironic if we discover the universe is just God's belly, and that astronomy is essentially cosmic gastronomy?

Healthcare analytics? Computational omics accelerate the life-sciences exploration metabolism ( ) Thurs #IBM quick-hit

RT @kexpplaylist #kexp Werewolves of London by Warren Zevon from Excitable Boy JK--I always wait for the last line: "His hair was perfect."

Roger Daltrey performs "Stand By Me"/"Wont Get Fooled Again" honoring Churchill bust at US Capitol. Wot? Beatles wont license "Hey Bulldog"?

It's better to focus on yourself originating stuff than on yourself originating original stuff. Better to pump it out, let others decide.

RT @kexpplaylist #kexp Night Time Is the Right Time by The Sonics JK--1965. Amazing! Only time I've heard somebody do a Ray better than Ray

RT @kexpplaylist #kexp Rock and Roll Halloween by Bobby Bare, Jr from A Storm - A Tree - My Mother's Head JK--My new favorite Halloween song

"Self-driving cars could save more than 21,700 lives, $450B a year" ( ) JK--Assuming they make them 100% roadworthy.

Drafted my next #IBMDataMag article: "Recommendations Galore: Exploring The Analytic-Powered Process Engines of the New Economy"

A8: #ibmblu What's next is #smarterplanet infra delivering right "signals" and giving confidence that we can ignore "noise" it filtered away

A8: #ibmblu What's next is more focus on collab human curation of new and existing data sources to facilitate tagging by "signal" relevance

A8: #ibmblu What's next is more adv cognitive computing techs that can spot unfamiliar signal in unfamiliar sources without pre-training

A8: #ibmblu What's next is more noise & signal, plus more automated infra baked into #bigdata #cloud #social #mobile #iot to distinguish

A7: #ibmblu Better sifting of signal from noise requires all users have access 2 self-svc, interactive, visual-rich #businessintelligence

A7: #ibmblu "Signal" delivery depends on tools that support contextual delivery of right data at right time to right decision maker.

A7: #ibmblu Distinguishing data signal from noise requires: #bigdata, stat analysis,profiling, filtering, quality, search, visualization.

A6: #ibmblu Augmentation of new signals (video, voice, etc.) benefits from correl8n w/existing data, eg, 360-degree customer profiles in EDW

A6: #ibmblu Deep metadata for new streaming sources can come in automated fashion from video/speech/image analytics tools.

A6: #ibmblu Can augment reading of new signals--voice, video, text, etc.--through crowdsourced or otherwise human-tagged metadata.

A5: #ibmblu Machine learning benefits from unsupervised and supervised methods of pattern-finding in new data sets.Both rely on horsepower

A5: #ibmblu Machines dont have severe attention, concurrency, & memory constraints o humans. Can multitask signal-finding at lightning speed

A5: About catching the bad data at the point of insertion/capture. The DBA needs to be able to focus on the client & not the SQL #ibmblu
 Retweeted by jameskobielus

#ibmblu A5. With BLU Accel, we see 20x speedup on identical hardware. Some queries over 1000x. Changes the scale of how data is explored.
 Retweeted by jameskobielus

A5: #ibmblu Plenty of algorithmic cognitive computing "sensemaking" functions--especially NLP--benefit from real-time hardware accel.

A4: #ibmblu DBAs now face challenge of optimizing queries, loads, job execution, & other functions across hybrid, distributed #bigdata infra

A4: #ibmblu The new business equation is balance need for #bigdata value against costs & complexity. Avoid overkill. Explore small data too

A4: #ibmblu The new #bigdata equation for business is find disruptive value in "signal" (eg microsegmentation) that's only evident at scale

A3: #ibmblu For my discussion of quality requirements re social sentiment data, see my #IBMDataMag article 

A3: #ibmblu Many operational data sets demand rigorous quality controls. But some sources, e.g., social intel, are less stringent re quality

Q3: #ibmblu Users must recognize that data, minus quality, is dangerous: can feed you w/bogus intelligence. Quality is key 4 decis support

A3: #ibmblu You dont "overcome" bad/extraneous data. Instead U profile, correlate, cleanse, correct, filter & augment 2 make it trustworthy

A2: #ibmblu Or construe "value" more business outcome that can be realized via evidence-driven analytics infused throughout biz processes

A2 #ibm Construe "value" as referring to any query that produces a data-driven response that you can actually trust & apply to biz question

A2: #ibmblu Analytics can drive value at any time & with any tool: simple BI query against an OLTP DB, or adv stat analytics on #bigdata

A1: #ibmblu Exploring the "noise" is heart of data science. Your data scientists use stat models & visual'zn 2 sift noise from the signal.

A1: #ibmblu Finding the "signal" sources isn't enough. Need to profile which data tables, records, & attributes are most relevant.

A1: #ibmblu What's in EDW may contain sufficient "signal." Or may need search, discovery, profilng tools 2 find "signal"-relevant sources

A1: #ibmblu You cut through noise by deciding what #bigdata sources are most relevant to your questions, and focusing on those.

RT @IBMbigdata: Make sure you check out background & #ibmibv study:  #ibmblu & this summ blog: 

Getting ready for #ibmblu Twitterchat "Distinguishing Signal Vs. Noise" at top of the hour. By "getting ready," I mean I have TweetDeck up

"Building models on seasonal data" ( ) JK--Discusses seasonality in performance (dependent var) vs. cohort (population)

"It's Your Data — But Others Making Billions Off It" ( ) JK--Compensate us when our social "recommendations" drive biz?

"Bigger Data? Or Better Models? Or….." ( ) JK--Not mutually exclusive. Bigger/better = larger samples & more attributes

"It’s everywhere! The day #Hadoop took over cloud" ( ) JK--Good provider quicklist. Incl. #IBM #BigInsights #SoftLayer

"Next Big Thing U Missed: #BigData Men Rewrite Govt’s Tired Economic Models" ( ) JK--What's with sexist "men" in head?

Sent my absentee ballot in for upcoming Virginia general election. Anybody who follows me knows I voted for McAuliffe (without enthusiasm).

@graemeknows You bet, sir! Thanks for stimulating my IOD-saturated cerebral cortex to the breaking point!

RT @HarvardBiz: Dilbert creator Scott Adams on management fads, The Office, & what his MBA taught him:  @Dilbert_Daily

Small data? Even no, null, and/or "naughty" data can yield valuabl insights p1  p2  Wed #IBM q-h

RT @kexpplaylist #kexp The Gift by Velvet Underground from White Light White Heat JK--Just realized it's "sworn vows," not "swollen vowels"

Catch me others Thurs Oct 31 12-1pm EDT #P4SPchat: Cognitive Computing ( )

Video playback: last Fri, Smithsonian "Surveillance Society" event I spoke at ( ). Me: 31:30-53:00 & 1:19:00-2:10:00

Lou Reed "Dirt" ( ) JK--1978. Perfect example of ballsiest Lou.

Lou Reed "Disco Mystic" ( ) JK--1979. This sounds like his "One Trick Pony" hack-producer persona kidnapped ballsy Lou

RT @kexpplaylist #kexp A Hazy Shade of Winter by S&G JK--Lou Reed played idiot record producer in "One Trick Pony" ( )

"The Myth of the War of the Worlds Panic" ( ) JK--I suspected as much. Newspaper people discrediting rival news medium

Come to Data Scientist Day, Mon Nov 4, #IBMIOD ( ). I moderate this expert panel:  w/Jake Porway

Autumn almost halfway over & our trees are still green & leafy. What gives? I recall northern Virginia falls that were colorful by Halloween

"#IBM Analytics Study Reveals #BigData Equals Big Payoff" ( ) JK--See overvu in my blog today ( )

RT @kexpplaylist #kexp Normal Person by Arcade Fire from Reflektor JK--2013. Good new. Ironically played while I read WashPost critic pan LP

Moose die-off in southernmost habitat due 2 heat stress frm global warming. Frostbite Falls MN still worthy of name? How's Bullwinkle doing?

RIP Marcia Wallace. Best known as voice of "Simpsons" teacher Edna Krabappel. I loved her as wisecracking receptionist on "Bob Newhart Show"

"Yes Twitter flawed during event like Boston bombings & so's everything else" ( ) JK--Cites #IBM social analytics study

New #IBM jk blog: "Converting #BigData And Analytics Insights Into Results: IBV Study Findings" ( )

"Why Facebook Is Teaching Its Machines to Think Like Humans" ( ) JK--"Deep learning" aka "cognitive computing."

"Convergence: HPC, #BigData & Enterprise Computing" ( ) JK--Interesting table of expected data volumes by data types.

Recommendation engines? Analytics to illuminate way out of stab in dark p1:  p2:  Tues #IBM q-h

RT @kexpplaylist #kexp Femme Fatale by Tracey Thorn from A Distant Shore JK--1982. Gr8 VU cover. Solo LP just before Everything But the Girl

Yesterday I Twitter-followed the Guggenheim, and now the Met, Whitney, & New Yorker following me. I must keep this riffraff at arm's length.

Many obits mentioned that "Walk on the Wild Side" was Lou Reed's only top-20 charting song. Funny that none dared call him "one-hit wonder"

Yoga practitioner is called a "yogi." I was shocked when I learned that yoga is Sanskrit word meaning "smarter than the average bear."

Saw "The Big Wedding" on DVD. Steaming pile of unadulterated Hollywood rom-com manure. Do De Niro Keaton Sarandon need the money that badly?

Drafted new #IBM jk blog: "Cognitive Computing: Relevant at all Speeds, Scales, and Scope of Thought"

"Man Who Would Teach Machines to Think" ( ) JK--Profile of Douglas Hofstadter. "Gödel, Escher, Bach" taught me to think

"Tuning Distributed Applications to Access #BigData" ( ) JK--Good practical guidance for DBAs.

Remember SNL Weekend Update several years ago when real Lou Reed responded 2 rumors of his death. "They're true. I've never felt more dead."

RT @kexpplaylist #kexp Romeo Had Juliette by Lou Reed from New York JK--Brilliant story-song, a la "Walk on the Wild Side."

"Cloud Predictive Analytics Most Used To Gain Customer Insight" ( ) JK--Presents interesting web user-survey results

Each of th 4 original Velvet Underground studio LPs was a classic, & quite stylistically distinct from each other & from 1967-70 rock scene.

A8: #cxo They used to say "an apple a day keeps the doctor away." It'll become "a healthcare app to guide your day keeps illness at bay."

@nvchawla A7: #cxo Clearly, the governance/quality of these records is a shared responsibilty of the collecting orgs (e.g., HMOs) & patients

A8: #cxo Consumers are already driving the personalized #healthcare revolution through smartphones, wearables, quantified life sensors, etc.

A7: #cxo Each consumer will insist on having full visibility w/introspection & ability to correct their medical profile.

A7: #cxo Mgmt of patient PII & EMRs is something I can speak on. Quantified life means each of us will own/manage what we self-collect

A7: #cxo IANAL (I am not a lawyer). The issue of patient PII ownership rights falls into category of topics I can't address knowledgeably

A6: #cxo Downside of personalized medicine is that may put too much reliance on patient self-svc, with doctors growing distant/inattentive

A6: #cxo Disad of personzd medicine is that physicians may focus so much on analytics that they fail to develop bedside manner/sensitivity

A6: #cxo Disad of personzd med is that patient & doctor may get so fixated on metrics that they overlook immeasurables (eg quality of life)

A6: #cxo A "downside" o personzd med is that we may acquire more insight into our unique profile than we would wish--eg., genetic diseases

A5: #cxo Personalized medicine means we're not being treated as specimens in the aggregate but as unique, complex individuals

RT @giasison: A5 Personalized medicine tailor fits your medical needs thus becomes more patient-centric and results oriented #cxo
 Retweeted by jameskobielus

A5: #cxo Personzd med also brings new self-svc model--patient can do/learn more on own--into doctor-patient relationship. Informed consumer

A5: #cxo Personzd medicine brings oldest "oldschool" doctor approach--direct nuanced 1:1 engagement--back in force in online-centric model

A4: #cxo Another big challenge in personzd medicine is encouraging physicians 2 change practices. Oldschool stubbornly proud of old ways

A4: #cxo Big challenge in personalized medicine is encouraging patient self-service through smartphone apps, portals, etc. Culture shift.

A4: #cxo Big challenge in personzd medicine is ensure privacy of patient PII. Another is analytic empowerment of smart choices by caregivers

@tmustacchio @NeilRaden #CXO What's a "pharma hack"? Article quality (wheat v. chaff) has diverse criteria. #Watson does smart search

A3: #cxo The doctor's expert judgment is paramount. Personalized treatment may eliminate unnecessary treatments if shows some ineffective

A2: #cxo Predict/prevent diseases requires that doctors examine patients directly & enter observed factors into comprehensive data profiles

A2: #cxo Personalizing care depends on 360-degree data view of each patient as an individual, incorporating histories, EMRs, outcomes, etc

A2: #cxo Health data on patients, their family histories, and communities helps to find complex trends and risk factors otherwise invisible

A1: #cxo #BigData helps doctors to aggregate patient history, correlate with current condition, & predict futures under various treatments

A1 Aggregate the risk factors then roll up into one, segregate then we usually see a pattern that may reveal rare diagnoses #cxo
 Retweeted by jameskobielus

A1: #cxo #BigData may help doctors model multlilayered physiological interactions that may have caused condition or are relevant 2 treatment

A1: #cxo #BigData may help doctors identify likely causes of conditions, and thereby suggest tests or treatments that may benefit.

A1: #cxo #BigData can uncover similar cases that doctors may no knowledge from their practice, or have overlooked in the medical literature

@IBMbigdata It's a pleasure to joining the #CXO today. Check out my relevant blog: 

RT @kexpplaylist #kexp All Tomorrow's Parties by Velvet Underground & Nico JK--German-Hungarian Nico sounds vaguely Irish on this: "mournin"

New #IBM jk #Dataversity column: "Machine Learning Boosts Data-Scientist Learning, and Vice Versa" ( )

"Made in #IBM Labs: Securing Data 4 Cloud Computing" ( ) JK--Patented invention secures priv info pre-transfer 2 cloud

"SoftLayer an #IBM Co & Cloudera Launch Turnkey #BigData Solns on Bare Metal Svrs" ( ) JK--SoftLayer design/deploy tool

Join #CXO Mon 12 ET! Personalizing #Healthcare Byte by Byte  w/ yours truly @nvchawla & @giasison #hcsm #bigdata

Mitch Ryder "Rock and Roll" ( ) JK--1971. I know it's sacrilege but I dig this version of th Lou Reed-penned VU classic

@nvchawla Thanks. Are you going to comp me on this? There's a paywall there.

#Hadoop uber-alles? YARN unwinds the grip of MapReduce uber-alles ( ) Monday #IBM quick-hit

RT @kexpplaylist #kexp Nobody but You by Lou Reed / John Cale from Songs for Drella JK--1990. Very cool & louche, tres Leonard Cohen.

RT @kexpplaylist #kexp Heroin by The Velvet Underground+Nico JK--Fun fact. Mayor Bloomberg finally got rid of dead bodies piled up in mounds

WashPost inexplicably prints most inane Lester Bangs diatribe against Lou Reed. Reminding me why I couldn't stomach Bangs' crappy criticism.

#IBM #BigData virtual event for #IBM India October 29/30. Keynote includes @jameskobielus, GPFS & #Hadoop, BigSQL  ^KS

Lou Reed's output was maddeningly bipolar and erratic from first to last. He took crazy artistic risks that mystified fans & critics alike.

RT @kexpplaylist #kexp Goodbye Old Friend by The Devil Makes Three from I'm A Stranger Here JK--2013. Beautiful. Sounds like Everly Bros gem

RT @kexpplaylist #kexp Temptation Inside Your Heart by The Velvet Underground from VU JK--One of their coolest not-originally-released.

Maybe I'm imagining, but #KEXP DJ Quilty 3000 sounds like she's holding back tears over Lou Reed's death. I've never heard that before.

RIP Lou Reed. One of the most influential artists in rock history, right up there with Elvis, Chuck Berry, Dylan, the Beatles, and Hendrix.

If you call yourself a "world citizen," does that mean you're volunteering for frontline duty when Mars attacks? Go right on ahead.

This Saturday I fly to Las Vegas for close to week's worth of nonstop activity at #IBMIOD. Every year, we collectively tense up to the event

Yoga. Relaxation exercises. It's all about finding ways to get even lazier deeper down in your very being.

RT @SI_Invention: .@jameskobielus: big data = extracting deep value from advanced analytics on trustworthy data at extreme scales.

RT @kexpplaylist #kexp Centaur by Buck 65 frm This Right Here Is Buck 65 JK--Funny/clever string of "I'm hung like a horse" double-entendres

Saw "Before Midnight" on DVD. Conversational saga o Jesse & Celine continues: 40-something, married, kids, & relationship issues. Still real

I noticed that 72-year-old Raquel Welch has titled her memoirs "Behind the Cleavage." I'll bet she's been waiting years to use that phrase.

RT @eliasisrael: To me, the phrase “self-driving car” sounds about as appealing as “self-eating food."

@allwiley My pleasure. I needed to teach the Smithsonian a few new words. Their vocabulary is so limited.