Monday, August 31, 2015

M is for Millenial?

Interesting piece on Facebook's new Messenger AI, "M":
When you ask M a question, the AI works to understand what you’re asking and formulates a response. But rather than sending it to you, the system sends this response to human “trainers”—customer-service types who work alongside the Wit.ai team inside Facebook’s new building in Menlo Park, California. These trainers then decide what else must be done to provide what you’re looking for

So it sounds like M gets the input, and then essentially send it to a Millenial, who Googles etc., to get the task done. I guess that works, and I think there are some startups taking this approach ("don't scale!") Fun to see it filter to FB, and a great cultural fit for them too. And now we know what "M" stands for.
In doing that heavy-lifting, the humans generate a roadmap for how particular questions should be answered. “Everything the trainers do, we record every step,” Lebrun says. This includes what websites they visit, what they say when calling the DMV, what they type in response to M users, and so on. In the future, this data can help drive a more advanced system based on deep learning, a form of AI that masters tasks by analyzing enormous quantities of information across a vast network of machines.

The long term plan is to gather inputs and automate the more common queries. Sounds great if it works. Siri has been a big disappointment as Apple's ability to deliver great experiences suffers post-Jobs.

Tuesday, August 25, 2015

Krugman out of paradigm

Although it's a terrible word, and an unhelpful communication technique, out of "paradigm" is still the best way I can characterize the continuing struggle academic economics has with understanding how finance, and particularly the monetary system, actually work.

Mosler highlights this when reviewing a recent Paul Krugman column:
Krugman: I know that may sound crazy. After all, we’ve spent much of the past five or six years in a state of fiscal panic, with all the Very Serious People declaring that we must slash deficits and reduce debt now now now or we’ll turn into Greece, Greece I tell you.

But the power of the deficit scolds was always a triumph of ideology over evidence, and a growing number of genuinely serious people — most recently Narayana Kocherlakota, the departing president of the Minneapolis Fed — are making the case that we need more, not less, government debt.

Why?


Mosler: This is the right answer- because the US public debt, for example, is nothing more than the dollars spent by the govt that haven’t yet been used to pay taxes. Those dollars constitute the net financial dollar assets of the global economy (net nominal savings), as actual cash, or dollar balances in bank accounts at the Federal Reserve Bank called reserve accounts and securities accounts. Functionally, it is not wrong to call these dollars the ‘monetary base’. And a growing economy that generates increasing quantities of unspent income likewise needs an increasing quantity of spending that exceeds income- private or public- for a growing output to get sold. 
Krugman: One answer is that issuing debt is a way to pay for useful things, and we should do more of that when the price is right.  
Mosler: Wrong answer. It’s never about ‘when the price is right’. It is always a political question regarding resource allocation between the public sector and private sector.
A lifetime ago, Krugman wrote that mathematical models were useful because they took implicit, inconsistent assumptions and make them both explicit and consistent. This was an aid to clear thinking.

His current thinking on monetary operations has a number of implicit assumptions, which is why he believes a fiat state has the same constraints and responsibilities as a household, and why his thinking fundamentally comes from the "sound finance" school of thought and not the "functional finance" school of thought proposed by Abba Lerner back in 1951.

Monday, August 24, 2015

Gell-man's law and Kuran

Gell-man's Amnesia Effect goes:
You open the newspaper to an article on some subject you know well. In Murray's case, physics. In mine, show business. You read the article and see the journalist has absolutely no understanding of either the facts or the issues. Often, the article is so wrong it actually presents the story backward—reversing cause and effect. I call these the "wet streets cause rain" stories. Paper's full of them.
From Michael Chrichton.

Kuran argues:
One mechanism [for authoritarian regimes in the Islamic world] is kinship ties. Because organizations could not form, people relied on their blood ties. This was part of the reason that commercial transaction were and still are (relatively) personal rather than the impersonal model generally followed in developed states. 
From a mugwump (via Andreeson)

Alternatively, it could be exactly the opposite.

Friday, August 14, 2015

China and the Yuan

If China's devaluing the Yuan, it suggests that the economy there is slowing down too much. China remains export driven, and a cheaper currency makes exports more competitive.
Sources told Reuters that the move to devalue the yuan reflects a growing clamor within Chinese government circles for a devaluation of perhaps up to 10 percent to help struggling exporters.
Fundamentally however, exports mean that a country generates real goods, and then exchanges them for numbers in a spreadsheet, overall it's a better deal for the importer than the exporter. Life in an export-driven economy means you work hard, but don't end up with much for it because most of your output has been traded away for digital beads. Mosler (and MMT) argues:
In a weakening global economy from a lack of demand (sales) and ‘western educated, monetarist, export led growth’ kids now in charge globally, the path of least resistance is a global race to the bottom to be ‘competitive’. And the alternative to currency depreciation, domestic wage cuts, tends to be less politically attractive, as the EU continues to demonstrate. 
The tool for currency depreciation is intervention in the FX markets, as China just did, after they tried ‘monetary easing’ which failed, of course. Japan did it via giving the nod to their pension funds and insurance companies to buy unswapped FX denominated securities, after they tried ‘monetary easing’ as well. 
The Euro zone did it by frightening China and other CB’s and global and domestic portfolio managers into selling their Euro reserves, by playing on their inflationary fears of ‘monetary easing’-negative rates and QE- they learned in school
No country is pursuing a fiscal strategy for increase demand and reduce unemployment. Ultimately, fiscal is all that will work, so we have our mix of deflationary low interest rates, increased risk of credit problems, and export-driven devaluation.

Wednesday, August 12, 2015

Is Alphanet like Berkshire Hathaway?

Google has rebranded itself as Alphabet. People are suggesting that Alphabet is like Berkshire Hathaway. I don't think that is an apt analogy, I think GE is a better analogy, but even that falls short.

Berkshire Hathaway is essentially a financial investing vehicle for Warren Buffet. He uses it to buy companies that are cheap. Alphabet is not fundamentally a financial investing vehicle, and it will not be used to buy companies that are cheap. While Buffet is folksy and charming and well loved, Alphabet will be making venture style investments not private equity style investments. Being the Berkshire Hathaway of the Internet sounds great, just like being the Google of the Beverage Industry, but I don't see how it goes further than that.

GE is a better analogy because it is a conglomerate, like Alphabet, with independently operating units, which share broad market information to work a little better. This is very Google, and it should be very Alphabet since Google has a strong competitive advantage in knowing almost everything that happens on the Internet and where the money is (via AdWords, Search, and Analytics).

GE also has world famous management discipline because running conglomerates is hard, and I'm not sure if Google has that discipline now but hopefully it may develop it in the future.