2013 December 28 Saturday
Washington State Credit Downgrade If 777X Production Leaves
If a machinists union does not vote to accept a retirements benefit cut then Boeing will likely move the new 777X elsewhere and Washington State will get a credit downgrade that will raise borrowing costs.
Washington faces a credit downgrade, higher borrowing costs and the loss of jobs and tax revenue should Boeing Co. (BA) decide to move production of its new 777X jetliner to another state.
Keep in mind that if Washington State had lower debt and spending then loss of production of a major airplane would not cut their credit rating. They've got to have other problems that put their creditworthiness in question for this to push them into a lower rating.
Meanwhile, socialist candidate Kshama Sawant, an economics teacher from India who accuses Boeing of "economic terrorism", was elected to the Seattle city council on a platform calling for an increase of minimum wage to $15 per hour. For the sake of the Seattle upper class I hope she can enact the high minimum wage. It will drive the more of the lower classes out of Seattle speed the gentrification (upper classes replacing lower classes) of the city. Hurray! We can have a one class society if we can just get the lower classes to leave. Yes, a one class society. That's where I want to live.
Should Boeing be deported from the United States, its factories moved to another country, so that the good people of Seattle can live free of economic terrorism? Why not even ban Boeing aircraft from the Seattle airport? Instead aircraft from socialist France (where 55% of GDP goes to the government - now that's socialist) could ferry workers to and from the soon-to-be socialist paradise Seattle.
Higher labor cost states have been losing manufacturing jobs faster than the cheaper states. For example, New York is losing manufacturing much faster than Florida.
New York state lost 42 percent of manufacturing jobs from 1990 through 2006. Over the same period, Florida lost 18 percent.
Cheaper states, other countries, and robots are all slashing the demand for high paid union factory labor. Robots will do the most of the manual labor. People with doctoral degrees are seeing the biggest increase in demand for their labor. Engineering and math doctorates are probably in most demand. Machine learning Ph.D.s baby. That's where the future is at.
By Randall Parker at 2013 December 28 06:45 PM
Until some variant on the Hutter Prize for Lossless Compression of Human Knowledge gets billions of dollars, I don't believe quality employees in machine learning are going to get hired simply because the money is being diverted by sycophants. They hiring is going to be directed toward kind of AI drek that was all over Minsky's "frames" back when people with any talent were trying to resurrect neural nets. Although, I'll admit, that drek is now all over neural nets -- 30 years late...
The increase in demand for Ph. D.'s is probably limited to STEM, and not all of them. On the downside, about three-fourths of all engineering and science Ph. D. candidates are foreign, and mostly Asian. These people go home. But if the immigration bill in the Senate passes, they will almost automatically get HB-1 visas and will stay. That will crash the labor market for STEM degrees.
We also have heavy over production of BS STEM degrees, and a substantial fraction of the recipients do not work in the discipline they trained in. It's about 50% for civil engineers.
If Boeing moves 777 production to another state, 747-8 production will go with it as both of these planes are manufacturing in the Everett facility. In other words, it looks like Boeing may be closing its Everett facility. This is a huge facility. I actually have no problem with a $15 minimum wage in Seattle area. However, a lot of restaurants and coffee shops will go out of business or will have to invest in automation. Automation is coming to the food service business anyways.
Large numbers of engineers are learning a large variety of ML techniques. You can watch Andrew Ng teach it on Youtube or watch a Coursera course. Companies in many industries are getting big ROI from ML specialists. ML is being used by biological researchers. I come across ML models in genetics, epidemiology, and other areas of biomedical research. There is no way the sycophants are showing up in so many places and blocking progress.
The many efforts on ML model development wouldn't be happening if there wasn't big ROI. Quality people in ML will bring in so much money that their businesses will grow. ML models succeed. If you aren't hearing about it you aren't at one of the right companies.
I just found Apache Mahout and am surprised to see it supports many things I hear from ML specialists such as Latent Dirichlet Allocation, a complementary naive Bayes classifier, K-Means, and other stuff.
Obviously the civil engineering students should have studied chemical engineering or even petroleum engineering (though I do not expect the demand for petroleum engineers to remain high). People definitely should steer clear of biology too and choose chemical engineering over chemistry.
I've no doubt that big returns are being achieved with machine learning techniques. Indeed, I haven't had a doubt that such would be the case since I took my first neural network course while living in La Jolla in the 80s and went on to build a 4e9 connections per second multi-source image segmentation system using Datacube's finite impulse response hardware in the late 80s -- the highest speed neural network ever built at the time. Keep in mind that a lot of these techniques people are calling "new" have been around for decades -- and that includes so-called "deep learning" which dates back to Hutter's colleague, Schmidhuber's work with recurrent nets in 1992. I'm currently working with the guy who financed the two-volume PDP books by Rumelhart Mcclelland and gave Werbos his first secure job (secure enough to raise his kids).
What I'm talking about has nothing to do with whether these sycophants are realizing actual ROI. I'm talking about the enormous gap between people who control the purse strings, and people who broke the real ground -- a gap being filled by people who are STILL decades behind and are the kind of "machine learning" intellects that would probably have been thinking Minsky was a genius back in the 80s.
When I talk about funding the Hutter Prize or similar as a touchstone I'm not merely indulging in some tunnel vision about machine learning -- some specialized benchmark that has little applicability to the vast array of technologies going by the name "machine learning". I'm talking about an operational benchmark for universal artificial intelligence that has very general applicability. Its more directly applicable to universal artificial intelligence than is the M-Prize directly applicable to SENS. The sycophants don't get it just like they didn't get neural networks back in the 80s. This kind of court sycophant has thus far successfully blocked attempts to fund it because, quite simply, they're as stupid when it comes to advanced theory as they are geniuses when it comes to playing court toadies.