Watson vs. Jennings

The best game show contestant on the best game show ever (Ken Jennings on Jeopardy) will soon compete against IBM’s Watson-class computer. Watson is favored to win. Today’s Watson technology requires a large room full of hardware. In 15 years, this same computing power will sit on a desktop. As we approach Turing-class computing (2050?), the power of today’s Watson computer will seem as quaint as the Sperry-Univac is today. I can envision a day when we no longer define religion in terms of having the “right knowledge.” I’m thinking David Hayward could do a cartoon showing three of the world’s leading religious thinkers as Jeopardy panelists – a Christian, a Muslim, a Buddhist – while Watson provides far deeper and more nuanced answers – even what we would call wisdom. And why not? With the entire written history of world religions (and every other branch of knowledge) in its memory (think Internet 2050), and thousands of increasingly nuanced processing algorithms at its disposal, it’s only a matter of time (perhaps 2 or 3 generations) before “epistebots” and “theobots” surpass the best human experts in their ability to parse and disseminate specific knowledge and even wisdom. Consider that this “entire body of human knowledge and wisdom” will one day be on every global desktop, in every language, in every home and mobile device, instantly accessible in deeply interactive and immersive ways we cannot imagine today. This is good news for religion, and humanity in general. As we are released from the need to define ourselves by tribal knowledge, we begin to define ourselves more relationally, more collectively, more empathically, more humanly. Even the perception of “enemy” must change as we begin to see ourselves as part of a larger global family. As we relax our desire and need for intellectual power, we can focus more on what makes us uniquely human:  our childlike awe and wonder at the universe and our ability to feel and act with greater compassion, empathy, and love towards other people, and all creation. “You have to understand all the nuances, all the regionalisms, slang and shorthand to play the game, to get the clues.” – Harry Friedman, Executive Producer,  

Looking

When I look at the ocean for a long time, the blue and restless driven waves, I keep looking, I keep looking, I keep looking at the waves swaying in the wind like a metronome, wired for the sound of a sleeping heart, and I keep looking with the silence of the sun on the windowpane, and I keep looking and do not stop looking deeper into waves as if into the middle of a woman’s body, where the soul and spirit have no human bonds, and I begin never to turn away from looking though I am frightened but keep looking beyond what I know until I can hardly think or breathe because I have arrived, with the need to be me disappearing into the beautiful waves, reflecting no one, nothing, no one. – Jason Shinder (1955 – 2008) …………………………………………… “All of man’s troubles stem from his inability to sit quietly alone in a room”  – Blaise  

The MAE Foundation

Gathering up those things which shape our thoughts We pack as though a journey to forever is stretching out ahead Saints go naked Prophets look back wistfully Choking vapours drive life out of corners And rivers graciously receive our poisons Like trusting children Patient figures stand and wait Gazing down at rusting rails To unknown places with no names – Ralph Steadman 1997 (from Plague and the Moonflowers) My talented friend Richard recently told me of a new philanthropic work to help Burmese genocide refugees (150,000+). The work employs music and musical instruments to enhance opportunity for creativity and hope into lives that survive on $20 per month. Richard has seen how music can positively change individuals and communities oppressed with struggle and little hope. Fittingly, the name of this new work is The MAE Foundation (Music Alters Everything). Richard’s vision includes the mobilization of the music industry (composers, producers, musicians, etc.) to become involved in this important work. I browsed some YouTube videos on Burmese genocide to put in this post, but literally broke down after a couple minutes. The inhumanity is unspeakable. I encourage you to browse over to Richard’s site (in beta), and do what you can to  

AD596

I’m particularly happy about a new product just announced at the 2010 San Francisco AES Conference (Audio Engineering Society). We call it the AD-596. It’s an 8-channel analog to digital converter of exceptionally high sonic performance. I’ve been using the AD-596 for critical listening tests in my lab and can confirm that it outperforms other well-known ADC designs at 2 or even 3 times its price point. Besides setting a dramatic new sonic value point in professional audio conversion, the AD-596 is also the world’s smallest 8-channel ADC, requiring just a single ‘500’ rack space (5.25″ x 1.75″). Up to 80 channels of this ultra-transparent converter can now fit in a single 3U 19″ rack. It was also confirmed by API that our AD596 is the first known digital-audio product for the 500-style rack. Some of the internal features include over-engineered AES transformers designed and built exclusively by Millennia, exceptional clocking circuitry of vanishingly low jitter performance in both internal and external modes, 90% efficient isolated switching power supply, ultra-quiet radiated and conducted performance for use with adjacent high-gain analog 500-rack preamplifiers, and premium components used throughout. Video report here:  

Hide and Seek

The PopTech folks have put Imogen Heap’s impromptu performance of Hide and Seek on line. She was not scheduled to attend the conference, so it was a real surprise. And it answers that nagging question:  can IH really hit a high B, or is it auto-tune? (yes, she hits the B). She mentioned that it was her first public performance of this song without a vocoder. I had a chance to chat with Imogen earlier in the day in one of the upstairs rooms of the Camden Maine opera house. She is a very lovely lady. Here’s her original (vocoder + autotune)  

Photos from Maine

Beautiful Camden Maine. Trees are still turning. Was a nice surprise to see Alex and Erwin McManus here at PopTech today. The three of us took some time off for exploring and photographing the area. Also had a surprise visit and performance from Imogen Heap just a moment ago. She sang Hide and Seek – gorgeous. Hope to spend some time with mastering engineer Bob Ludwig tonight in Portland. Enjoy the  

POW-r Algorithms

Twelve years ago, some friends and I got together with the intent of developing the most musically neutral and dynamically accurate audio bit length reduction algorithms. As we completed the code, many of the audio industry’s golden-eared engineers and producers reviewed our work favorably. We soon after became the world’s #1 software for audio bit length reduction. The software is called POW-r, which is an acronym for “psychoacoustically optimized word-length reduction.” Most professional audio recording today uses DAWs, PC-based “digital audio workstations.”  Digitized audio is stored in software bit chunks called “words.” Most DAWs today default to 24-bit word lengths (although internal processing may be twice that or more). Each bit represents a 6dB change in “audio voltage.” More bits equals higher acoustic dynamic range. A higher dynamic range equates to more realistic sound reproduction. The common CD stores digital audio in 16-bit word lengths. And this is the problem: when transferring native 24-bit audio from the DAW onto a 16-bit CD, we lose 8-bits, or 48dB! What does 48dB sound like? It’s the difference between normal conversation (65dB) and a live rock concert (115dB), or the difference between a softly played piano (75dB) and a forte symphony orchestra (120dB). How do you get the full impact of a 24-bit studio recording (potentially 144dB*) onto a CD which can only represent 96dB? Enter the unique software algorithms called POW-r. Our code was created in the real world of symphony orchestras, of which I have engineered hundreds of recordings. We tested numerous iterations of the software in real-world acoustics, carefully comparing musical results until we found optimal subjective performance. Today, POW-r remains the world’s #1 word-length-reduction solution, both for CD and MP3 bit preparation. Most of the top DAW companies license POW-r (Apple Logic, Avid ProTools, Cakewalk Sonar, Magix Samplitude and Sequoia, Ableton Live, Pyramix, and many others). It’s been estimated that POW-r is now used on over 400 million CDs and downloads annually. (* in practice, studio recordings rarely achieve 144dB dynamic range, and home playback systems can rarely offer much more than 110dB, if that. What’s worse, most music today is played back into ear buds, with a dynamic range rarely exceeding 90dB, and that assumes a very quiet environment and high quality playback  

Harvest

It’s that time again – grape harvest in the Sierra Foothills of California. Unfortunately, we had one of the coolest summers on record so our own grapes (grown at nearly 3,000 ft. elevation) are unripe, and unlikely to ripen before frost. We purchased three varieties this year from other growers. My crusher motor stopped working, so some of these grapes were crushed under a pair of very large feet – size 14 (Keen) or 15 (Nike). 500 pounds Grenache 300 pounds Pinotage (a South African variety, rare in the USA) 600 pounds Syrah This year, I’m fermenting in small (100 pound) batches in food-grade bins. This gives me the opportunity to experiment with different kinds of yeasts on the same grape and taste the results before choosing to barrel blend. It also allows increased skin-to-juice contact during maceration. Apparently, there are only 40 acres of Pintoage planted in the USA, and half of those acres are here in the Sierra Foothills. My friend Aaron has been making a Pintoage for years, and it is always exceptional. This will be my first experience with the grape, which is a genetic cross of Pinot Noir (Burgundy) and Cinsault (Rhone). When we picked the Pinotage, sugar content was at 30 brix. This is a MASSIVE amount of grape sugar, but the ripeness gave the grapes a really deep and rich flavor, almost raisin-like. At 30 brix, no yeast I’m aware of can handle the eventual alcohol (18%), so after a two-day cold enzyme soak, I diluted the must to 25 brix and started fermentation with slow starting, slow working Laffort FX-10 yeast. It’s said that Pinotage delivers better fruit flavors with a slow fermentation. The Grenache came from my friend and neighbor John, coming in around 24 brix with high TA (0.9). I drew off 6 gallons of early crush, which will become a Grenache Rose. Am trying a new yeast called Lalvin Rhone 4600, which promises to bring out “complex aromatic notes and elevated ester produc­tion such as tropical (pineapple) and fresh fruit (apple, pear, strawberry).” We’ll see… The color of this year’s Grenache is medium salmon. It will probably become blending stock for some other wines currently in barrel. The Syrah ended its cold soak last night. I’m experimenting with three different yeasts: D-254, D-21, and FX-10.  Am also separating for later blending decision the hard press from the free run, as each gives a unique character to the