Sunday, October 09, 2011

Siri, the Friendly AI

The iPhone 4S video shows a young runner asking Siri to rearrange this schedule. It doesn't show him running into the path of another Siri user driving his convertible.

Siri is the iPhone AI that understands how your phone works and, in theory, understands a domain constrained form of natural language. It has a long AI legacy; it's a spinoff from SRI Artificial Intelligence Center and the DARPA CALO project.

When Siri needs to know about the world it talks with Wolfram Alpha. That's where the story becomes a Jobsian fusion of the personal and the technical, and Siri's backstory becomes a bit ... unbelievable.

Siri was launched as the unchallenged king of technology lay dying. The Wolfram part of Siri began when Jobs was in exile ...

Wolfram Blog : Steve Jobs: A Few Memories

I first met Steve Jobs in 1987, when he was quietly building his first NeXT computer, and I was quietly building the first version of Mathematica. A mutual friend had made the introduction, and Steve Jobs wasted no time in saying that he was planning to make the definitive computer for higher education, and he wanted Mathematica to be part of it...

Over the months after our first meeting, I had all sorts of interactions with Steve aboutMathematica. Actually, it wasn’t yet called Mathematica then, and one of the big topics of discussion was what it should be called. At first it had been Omega (yes, like Alpha) and later PolyMath. Steve thought those were lousy names. I gave him lists of names I’d considered, and pressed him for his suggestions. For a while he wouldn’t suggest anything. But then one day he said to me: “You should call it Mathematica”...

... In June 1988 we were ready to release Mathematica. But NeXT had not yet released its computer, Steve Jobs was rarely seen in public, and speculation about what NeXT was up to had become quite intense. So when Steve Jobs agreed that he would appear at our product announcement, it was a huge thing for us.

He gave a lovely talk, discussing how he expected more and more fields to become computational, and to need the services of algorithms and of Mathematica. It was a very clean statement of a vision which has indeed worked out as he predicted....

A while later, the NeXT was duly released, and a copy of Mathematica was bundled with every computer...

... I think Mathematica may hold the distinction of having been the only major software system available at launch on every single computer that Steve Jobs created since 1988. Of course, that’s often led to highly secretive emergency Mathematica porting projects—culminating a couple of times in Theo Gray demoing the results in Steve Jobs’s keynote speeches.

... tragically, his greatest contribution to my latest life project—Wolfram|Alpha—happened just yesterday: the announcement that Wolfram|Alpha will be used in Siri on the iPhone 4S...

Siri's backstory is a good example of how you can distinguish truth from quality literature. Literature is more believable.

Siri isn't new of course. We've been in the post-AI world since Google displaced Alta Vista in the 1990s. Probably longer.

What's new is a classic Jobs move; the last Jobs move made during his lifetime. It's usually forgotten that Apple did not invent the MP3 player. They were quite late to the market they transformed. Similarly, but on a bigger and longer scale, personalized AIs have been with us for years.  AskJeeves was doing (feeble) natural language queries in the 1990s. So Siri is not the first.

She probably won't even work that well a while. Many of Apple's keynote foci take years to truly work (iChat, Facetime, etc). Eventually though, Siri will work. She and her kin will engage in the complexity wars humans can't manage, perhaps including our options bets. Because history can't resist a story, Siri will be remembered as the first of her kind.

Even her children will see it that way.

Update 10/12/11: Wolfram did a keynote address on 9/26 in which he hinted at the Siri connection to Wolfram Alpha: "It feels like Mathematica is really coming of age. It’s in just the right place at the right time. And it’s making possible some fundamentally new and profoundly powerful things. Like Wolfram|Alpha, and CDF, and yet other things that we’ll have coming over the next year." The address gives some insight into the world of the ubiquitous AI. (No real hits on that string as of 10/12/11. That will change.)

OWS Recruiting poster by Hargrave Yachts

Via G+. An obvious Occupy Wall Street recruitment poster has been cleverly planted in Yachts International Magazine as a (flash) ad for a Hargrave mega yacht.
Since I wonder how long the ad will actually last online, here's a screenshot and some copy ...

Screen shot 2011 10 09 at 5 20 14 PM

"We used to sell yachts as luxury items, in todays' world they're really a necessity.

Years ago buying .. the 136' Hargrave "DREAMER" ... was ... a once in a lifetime reward for those of you who made sacrifices ...

... Today things have changed ... successful people have now become the target of every two bit politician from the White House on down ...

... being on the water has become a necessity for today's entrepreneur ... a refuge where they can leave the inanity of today's world ..."

On the water, far away from the peasants.

Diabolically clever OWS. I'd never have figured Hargrave as one of your covert operations.

[Note: By a freak accident I managed to completely obliterate this post, but I was able to restore it.]

Why rationalists like OWS

Philosophically I'm closer to OWS than to the Tea Party. In terms of policy, I'm reluctantly closer to Clinton/Gore than to anarcho-socialism.

So why am I an OWS supporter?

I think I can put it into a picture like this ... [1]

AmericanVolume 2

The Y axis is volume in the American discourse; that's what enables political "courage". The X axis represent a range of governing policies. The blue box is the limits of the rational -- the range of policies that are most likely to lead to a relatively long and happy life for human civilization, democracy, and us.

Obama is to the left of that range (I'm deliberately flipping this with the conventional left/right picture of American politics.).

In the middle is GDKK - Gordon (hey, it's my blog), DeLong, Klein and Krugman -- in rough alignment with my TP to OWS spectrum.

At the right is Occupy Wall Street.

Without OWS the rationalists are at the far right of this X axis, and the "middle" of the American discussion is outside the bounds of reason. In this context a Carbon tax, for example, is inconceivably radical.

With a healthy OWS movement the volume shifts.

I think that will help.

[1] Google's drawing tool app is amazing. It's the best part of the Google Apps suite at the moment, which is saying quite a bit.

See also:

Saturday, October 08, 2011

Comparing carriers and networks for iPhone owners: AT&T, Verizon and Sprint

Reading this Marco Arment post, I wonder if I should have gotten Emily the 64GB model. I didn't think about what it meant to capture HD video. More than a phone discussion though, it's a carrier comparison. There's a lot I didn't know (emphases mine) ...

Which iPhone 4S should I get? – Marco.org

... Under ideal conditions, AT&T has the best data speeds and Verizon has the best voice quality, but almost nowhere actually provides ideal conditions, so it really depends on which carrier sucks the least in your area.

All three carriers’ plans are priced in the same ballpark.

People often say that Verizon covers fringe areas better, but in my experience, AT&T is comparable — sometimes one works and the other doesn’t, but neither more often than the other.

I don’t have much experience with Sprint, but the little I’ve had suggests that it doesn’t quite provide the coverage and strength that Verizon and AT&T do. Sprint phones can roam on Verizon’s network if there’s absolutely no Sprint signal, but in practice, that doesn’t happen often, and the phones will prefer a weak Sprint signal to a strong Verizon signal.

Sprint is the only carrier offering “unlimited” data.

Verizon messes with your data uncomfortably — they recompress JPEGs to save bandwidth, and they watch the sites you visit to collect (and presumably sell) the aggregate stats.

... any iPhone 4S you buy in the U.S. with a contract at a subsidized price (less than $649) is still carrier-locked. You can’t change carriers later...

The carrier lock really sucks. Even after you've paid your $800 or so for your $650 phone, it's still carrier locked. Of course it's two years old by then, but Emiiy's 2 yo 3GS still works real fine. Yet another reason that American geeks hate mobile phone companies.

If you're trying to find out which carriers work in your area, my friend Robert M recommends Antenna Search as a guide to your local towers. I discovered my home is in an area with relatively few antennae -- except for 3 immediately nearby. We don't have coverage problems.

Agriculture and population: which came first?

We're used to thinking that human population (and human misery) boomed when we switched to agriculture. Now it appears that a post-glacial population boom may have inspired agriculture ...

Dienekes’ Anthropology Blog: Major population expansion in mtDNA of East Asians

... the further analysis showed that the population expansion in East Asia started at 13 kya and lasted until 4 kya. The results suggest that the population growth in East Asia constituted a need for the introduction of agriculture and might be one of the driving forces that led to the further development of agriculture...

The crash of 08: an 8.9% collapse in GDP

The Great Recession, or, as I prefer, the Lesser Depression, included one of the worst quarters in American history ...

Could this time have been different? - Ezra Klein - The Washington Post

... The Bureau of Economic Analysis, the agency charged with measuring the size and growth of the U.S. economy, initially projected that the economy shrank at an annual rate of 3.8 percent in the last quarter of 2008. Months later, the bureau almost doubled that estimate, saying the number was 6.2 percent. Then it was revised to 6.3 percent. But it wasn’t until this year that the actual number was revealed: 8.9 percent. That makes it one of the worst quarters in American history....

The M word

We commies have a problem.

We are dependent on the money of corporations and the votes of the masses. It's not our only problem. We have to hold together a coalition across tribe and culture.

Since they're dedicated to the end of civilization, it's a good thing the GOP has problems too. Gopers have to sign up for a set of mandatory untruths on science, history and economics for example. This makes it hard for the GOP to retain non-sociopathic rationals, much less secular humanists.

That's not their big problem though. Their big problem is theological. A party famous for intolerance has to pretend two quite different religions, Mormonism and Christianity, are the same. (Trust me, I'm an atheist. These religions have more in common than Islam and Judaism -- but not much more.)

This is a lot to pretend, but it doesn't seem as hard as, for example, worshipping both Christ and Mammon. It would be doable -- except for the Hell problem. Many GOP Christians believe in Hell, and they believe consorting with heretics is one way to get there. Mormonism is heresy; no doubt there. An eternity in Hell is a high price to pay for winning an election.

Maybe the GOP will patch this up -- but I think they're going to have to choose. They'll either have to offend Mormons and Romney supporters or offend their entire Christian Base.

Despite his current dominance, I don't think Romney will be the GOP candidate. If he is, he'll have a hard time harnessing the base.

If Perry is the alternative, expect Obama to ask if he'd have been welcome at Perry's *head ranch.

Considering the horrible odds against Obama's reelection he'll need more of this Irish luck.

Anonymous

Google is trying to enforce full transparency in their, large, corner of the web. I think they're making a terrible mistake.

I can see why they do it though. Most large sites can't handle the spam attacks routed through anonymous posting.

Gordon's Notes, though, we're not so big. Most of the comments I get are anonymous (excepting Martin and MaysonicWrites). Many of them are excellent. Fortunately Google's AI is now pretty good at killing the spam attacks, so I can easily manage the low volumes I get. [1].

Anonymity is something a *cough* specialty blog like GN can support.

[1] Sure would help if Google gave me an RSS feed of comment counts though - so i know there are new ones to inspect. As it is comments go immediately to recent posts and I get a notice by email, but older post comments can sit around until I notice and authorize them.

99%

Of course I support Krugman's army. Not because they have answers; but because humans are the neurons of the mass mind. OWS is thinking in action. If nothing else, it's a start on a very long struggle.

Klein, Salmon and Collins have done a good job on the scene. Klein suggests the 1% may be feeling some heat ...

A breakdown of trust - Ezra Klein - The Washington Post

At a medical-innovation conference in Cleveland this week, I heard chief executive officers complain about being treated like “criminals,” about profit being regarded as intrinsically suspicious, about the president saying unkind things about oil companies, about exemplars of hard work and success being viewed as greedy rather than productive. Like Rodney Dangerfield, the rich may be making money, but what they really want is respect.

Even worse, they said, was the fact that their every suggestion for getting the economy back on track was denounced as a self-serving grab for more profits. Shouldn’t it be obvious to everyone that it’s better to give corporate America a tax holiday for overseas income, because otherwise the money will never return home? Isn’t it clear that we need more high-skills visas for foreigners? Have we seen the airports in Asia? Doesn’t the U.S. realize that if Germany and China and South Korea roll out the red carpet for companies, giving them every tax advantage and regulatory break to locate there, that the U.S. must up the ante?

The CEOs have a point. Not on the tax holiday for overseas income -- that’s a scam. But the U.S. could make it easier to do business here. We do need more high-skills visas. We do need to reform our tax code, reduce our deficit, upgrade our education system and repair our infrastructure. We even need to compete with the incentives these companies receive to relocate their factories and research centers; it’s a fact of the modern economy, and we can’t pretend otherwise.

But the self-pitying, self-righteous tone of these complaints misses the big picture, and makes the underlying problem worse: The rest of America doesn’t trust corporate America right now. The rich have been getting fabulously richer, corporate America is sitting on trillions in cash reserves, and where has that gotten the rest of the country? A shabby, jobless recovery in the early Aughts, followed by a credit bubble, followed by a crash in which ordinary Americans had to bail out Wall Street, followed by the worst economy in generations...

Krugman has said the something similar -- the OWS gang can't be as dangerously incompetent as the titans of industry and Wall Street. My recent post on the crummy option for middle-class investing is partly a reflection of this global mistrust. Why should we trust the financial statements publicly traded companies are producing? And if we can't trust those statements, how can we trust stock funds?

Good job OWS and OccupyMN. Keep at it.

The rest of us don't have to merely stand and wait. Even if we can't physically join OWS, or even local movements like Occupy Minnesota, we can donate though. Start with $25. (PS. The donation worked without providing a phone number of email address. The inevitable paper spam will at least help our ailing USPS.)

Friday, October 07, 2011

Investment in a whitewater world

During the last half of the 20th century retail investors earned positive returns with some mixture of stocks, bonds, real estate (personal) and cash. Mutual funds, and especially index mutual funds, made middle class investing possible.

Then came the great market bubble of the 90s, the real estate bubbles of the 00s, and the rise of IT enabled economic predation. The vast flow of growth returns was diverted from the middle class investor to corporate executives and sharper, faster, players. Corporate financial statements became less and less credible as new ways were found to obfuscate financial status. In a world of IT enabled complexity, risk assessment became extraordinarily difficult. Where growth opportunities were strong, as in China, the markets were corrupt and inaccessible.

Maybe we'll return to the relative calm of the 1980s, or even the slow growth of the 1970s. Oil prices may stabilize around $150/barrel. China's economy may make a soft landing and the Chinese nation may follow Taiwan's path to democracy. The GOP may shift away from the Tea Party base and, once in power, raise taxes substantially while implementing neo-Keynesian policies by another name. North Korea may go quietly. The EU may hold together while gradually abandoning the Euro. Cultural shifts might make personal integrity a core value. Disruptive innovations, like high performance robotics and widespread AI, may slow. The invisible hand and social adaptation may solve the mass disability problem of the wealthy nations. Immigration policy, a breakthrough in the prevention of dementia, or the return of ubasute may offset the impacts of age demographics. We may even ... we may even look intelligently at the costs of health care and education and manage both of them.

If these things happen then some economic growth will return, stock prices will reflect fundamentals, bonds will become feasible investments, and interest rates will be non-zero.

Or they won't happen.

In which case, we can look forward to more of the same. The best guide to the near future, after all, is the near past. In this whitewater world then, in which financial statements are unreliable and wealth streams are diverted, what are the investment opportunities? We cannot recover the lost returns of the past 11 years, but it would be nice to be less of a chump.

Real estate seems a reasonable option, though there we face the problem of untrustworthy investment agents. There is not yet a John Bogle of 21st century real estate investment. (This, incidentally, suggests something government could do -- engineer a trustworthy investment representative for American real estate.)

The other option is to switch from prey to predator.

During the 20th century retail investors could only make one way bets. We basically had to bet on economic growth and prosperity. For a time we could shift a bit. If we thought near term growth looked bad, we could shift to bonds. If we thought a crash was coming, we could shift to cash. Basically, however, we could only get good returns by investing in stocks and betting on growth. This worked under conditions of economic growth and relatively integrity. Under the conditions of the past decade this made us prey.

Predators don't make one way bets. They make bets on downturns, on upturns, on volatility, on stability, on irrationality, on continued fraud, on reform. They make bets on bets. They play the options and straddle options games that brought down the world economy. It's too bad they won, but, with a bit of help from our AI friends, they did.

So I'm learning about options. It's not what I like to do, but I don't make the rules.

Thursday, October 06, 2011

Saletan on the ethics of wealth and transplants

I'm a physician. I bet most physicians have been thinking about this over the past two years as Steve Jobs' health deteriorated ...
Steve Jobs liver transplant: Organ donation is the best way to honor him. - William Saletan - Slate Magazine

... there’s something you can do to help people such as Jobs. You can supply replacement parts for the machines that keep them alive. You can sign up as an organ donor...

... To get the liver, Jobs went to Tennessee, because the waiting list in Northern California was too long. There weren’t enough livers to go around. Lots of other people in Northern California needed livers but couldn’t get them, because they didn’t have the kind of money or savvy Jobs did. They couldn’t afford to fly around the country, go through extensive evaluations at multiple transplant centers, and guarantee their availability within an hour for the next liver that became available...

... Earlier this year, when Jobs took a leave from Apple because of deteriorating health, I asked whether he should have received his transplant in the first place. As bioethicist Arthur Caplan has noted, almost none of the 1,500 people who received liver transplants in the U.S. when Jobs did, in the first quarter of 2009, had cancer. That’s because there’s no evidence that transplants stop metastatic cancer. The much more likely scenario is that the cancer continues to spread and soon kills the patient, destroying a liver that could have kept someone else alive for many years. Among liver recipients, cancer patients have the worst survival rate. While more than 70 percent of liver recipients in Jobs’ age bracket are still alive and functioning five years later, Jobs lasted only half that long.
In theory wealth and genius are not part of the criteria for liver transplant. In theory, even goodness isn't part of the criteria.

In reality, money makes a huge difference with most things in life. Nobody has ever said Jobs was saintly. Most people with money and/or fame or both would have done what Jobs did. It would have been good if he'd left some money for less privileged people in need of new parts, but that wasn't his style.

Even though Jobs only did what was normal for the powerful, The Tennessee transplant center did worse than most. They do deserve some hard questions.

Wednesday, October 05, 2011

Jobs parents

I've read the obits. Steven Levy (Wired) and John Markoff of the NYT were best. The Economist disappointed.

I've also read many of the posts from longtime Apple bloggers. They rightly express sympathy to Jobs wife and children.

Noone, though, has expressed sympathy for Jobs parents - Paul and Clara Jobs. It is a terrible thing to outlive your child. From the bit I've read they gave their challenging son all they had. In a way they could not have imagined, their sacrifices changed our world for the better.

I assume Jobs did well by them.

My thoughts go to you Paul and Clara.

Update: Paul Jobs died in 1993 at age 70; Clara Jobs (nee Hagopian) died in 1986 at age 62. They both lived to see Jobs early success. I found one 9/2/11 article about them.

As the father of 3 adopted children I noticed a few things in the coverage of Jobs death. There is much more discussion of his birth parents than his adoptive parents. From what I have read, Jobs had little or no contact with or interest in his birth parents. I saw some articles claiming they might have some claim on Jobs estate. I'd be astonished if that were true.

I also notice that references to Jobs' adopted sister Patti seem to assume they were loosely connected. Jobs grew up with Patti; she was his sister.

Jobs parents seemed to value their privacy more than most. Jobs did too.

GTD: Create appointments not tasks

Since nobody will make the software I want I often have to choose between creating a task and creating a calendar appointment. Sometimes I do both, but mostly I do one or the other.

If it's a task I really need to get done (A) I schedule it on my calendar. I don't create a task. When I schedule it I add a note on what I want to do and what the next steps are (if any). I schedule the time I need to do the work. Sometimes there are some related tasks, but I try not to get fancy.

B Tasks are things I want to do. Sometimes I need to do them, but I may not have any capacity to schedule them. I put B tasks in my task/todo software and I usually assign them a target date based on their size and my predicted capacity. If they get more urgent I make them A tasks and schedule them (so in this case I have both a task and appointment, but it started as a task).

C Tasks are often ideas or future projects. Sometimes they're important, but can't be scheduled (too big, too expensive, etc). C tasks are a "backlog" of work. Periodically I purge them if they don't get done, but they mostly sit quietly in the queue, not bothering anyone. A C task can be promoted to a B task -- and it might even get a date. Or, if time/money arrives, they go straight to A.

See also:

Your public Facebook posts - try this Google search

If you've ever used Facebook, log out of Facebook then try this search:

site:facebook.com "your name"

I don't share publicly - but I do post and comment on "Pages" which belong to organizations.Those pages are always public, so what I have written there is also public.

You can't make these posts non-public, but you can delete them. Log in to Facebook, then repeat the search. You should now see a delete box.

I found some posts could not be deleted. I got the "failed to hide minifeed story" bug on one.

PS. In the midst of this exercise my (true and unusual) name was registered as a Tidbits author with a 1996 article. This was a puzzling experience, because it was at first completely unfamiliar. As I read it, however, it became vaguely familiar. I remember the ideas, if not the article. I'm pretty sure it is mine. Weird.

GTD: Introduction to a series

With this post I'm starting a 'gtd' labeled series of posts. GTD is an acronym for "getting things done", a productivity methodology that was fashionable a few years ago. It's less fashionable now, but it will return under another name. After all, the base approaches have been rebranded many times over the past two thousand years.

These will be short posts - shorter than this introduction and much shorter than my old GTD posts [1]. I'll space them out. If you're interested you can experiment and add things to your own way of working. If you're not interested, they should be easy to ignore. There won't be anything truly novel - this ground has been well worked.

Posts will focus on things I've done since at least 2004. I'll mention the tools I use, but the main focus will be methodology. Each post will be limited to one component spanning tasks/projects, calendaring, and email. They will incorporate my more recent experiences with Agile development methodologies.

See also: [1]