George Rebane

Thought I’d try something different on this cloudy and wet Saturday morning.  The Mueller report is in AG Barr’s hands, and from the early assessments it is a nothing burger for the TDS Dems.  If so, they are looking at it merely as a speed bump on the road to Trump’s impeachment, which we are assured will follow when the next tranche of “clear evidence” of some, any, wrongdoing will be dug up by the Left’s congressional witch hunts that have yet to get their second wind.  I’m going to hold back my sage assessment of the progress toward nailing the president before the 2020 election.

That China’s going to eat our lunch seems like a much more relevant issue to dwell on.  So argues AI expert and prominent VC Kai-Fu Lee in his AI Super-Powers – China, Silicon Valley, and the New World Order (2019).  Lee, a Chinese national, has impressive credentials – Google’s speech recognition head, then chief of Google China, then major VC funding AI start-ups here and in China.  He goes into great detail outlining our two countries’ different approaches to AI/products development and marketing, and the mentality of the entrepreneurs involved.  The punchline is that China has caught up or soon will in all the major arenas of commercializing and researching AI.

A careful read between the lines, makes me think that Lee really wrote this book for China’s communist leadership, complete with data-supported observations, comparative analysis, and critiques of both countries’ government support of their AI sectors.  He makes a case for state-controlled support of China’s AI, while carefully outlining the weaknesses of lawless competition and a command research environment that’s not all that conducive for creativity.  He ends the essay with a very hopeful vision of how AI-induced systemic unemployment may be alleviated (yes, Lee flies in tight formation with Rebane Doctrine on several important fronts).  The only big hole that I saw that he ignores the fact that half of our workforces have 2-digit IQs, and therefore cannot be ‘upskilled’ to fill future jobs.  Lee sees them providing the human touch and delivering “love” in human2human caregiving work.  No matter, for those interested in the last great century of Man, this is a worthy read.

Speaking of venture capital and private equity, there’s so much of it now sloshing around that portfolio managers have a hard time finding worthwhile projects to fund.  This means with all those bucks chasing promising projects, the expected return from such funds is going to suffer.  As an example, CA’s underfunded CalPERS govt employee pension fund has $358B to invest and is considering where to add to its $28B already in private equity, that in a desperate attempt to deliver on its la-la land promise to garner a 7% annual return so as to avoid a giant unfunded liabilities mammary in the wringer.  To date, it has yet to reach that level even for a single year, and now we have over $1.3T wanting to play in that sandbox.  That future doesn’t look good, especially if their darling Dems get their hands on the US economy – fugeddaboutit.  BTW, the teachers’ pension fund CalSTRS isn’t doing any better.

This, of course, reminds me of where AOC and her double dummies are planning on taking the country.  These vacuum-heads are champions of the liberal tenet that tax rates don’t affect investment risk – think Green New Deal and Medicare For All.  The entire notion of monetary utility is unknown to them.  Would you wager $100 on a 3:1 bet (Pr(win) = 0.75)?  Of course you would, and you’d play that game all day long.  Would you take the same bet for $500K – one out of four you’d be out of a half million?  I hope you’re getting the idea.  But we could come up with some threshold odds where you’d be willing to wage $500K.  Raising tax rates messes with the payoff, which lowers the odds.  Where do you think CA would be when Silicon Valley's odds finally dip below the risk tolerance threshold?  And that Calexit process is already under way.

Notice all the recent soul-searching articles on biased algorithms?  Most of the writers (especially if they’re journalists) don’t have a clue about bias or algos; their intent is to write the usual PC pap that the lamestream laps up and delivers.  The topic is vast, but can and should be approached with a few principles in mind.  First and foremost is that bias infers something that is unfair (and perhaps even unjust, but that’s another dimension – things can be unfair but not unjust).  The notion of achieving fairness among multiple parties ranges between very difficult and impossible as shown long ago by nobelist Kenneth Arrow.

So now we have an algo, say, used in social work to determine whether a battered kid should be returned to his parents.  The algo is discovered to be biased in that it denies such reunions more with parents of color than whites, on the basis that the kid would most likely suffer again.  So should the algo be pulled because it happened to be trained with data that contained a preponderance of non-white cases, no matter that it did a very good job predicting the safety of kids with colored parents.  So should we pull the algo from use and go back to the old way with unaided social workers making decisions that demonstrably has resulted in increased pain and suffering by colored kids, in order not to be biased?  Or should we continue using it because it clearly saves a known number children from being battered again by their mostly colored parents even though the reliability of such predictions aren’t as good with whites?

As a lifelong practicing decision theorist, the answer to that question is easy to prescribe.  What is the quantifiable social utility (remember, all utilities are subjective) that the responsible agency is attempting to extremize (max or min) through its prudent placement of battered children?  If it is to minimize the expected number of re-battered kids, then keep using the biased algo (while getting more white-oriented training data).  If it is to minimize the political (existential?) cost to department head careers from unsubstantiated charges of political incorrectness, then dump the algo.  Throughout my years, the understanding of this nuance has pretty much determined the ideological cast of the decision maker – hint to the reader: progressives don’t like quantifiable utilities, in fact they don’t like utility-based decisions in general – too restrictive.

More later.

[later]  Well, maybe a smidgen on Mueller.  Will anyone here or across the land now admit to the countless lies told by almost every Dem and lamestream commentator about the existence of “clear evidence” of “conspiracy” and “traitorous collusion” that supposedly threw the election in Trump’s favor?  (FN is having a blast playing archived videos of CNN, MNBC, …, and more Dem politicians than you can count, all blasting Trump with non-substantiated charges for two years, and they’re not done yet.)  And these are substantive lies, not the inconsequential bluster bullshit that liberals count as Trump’s lies.  Also take into account the alleged attacks on Mueller and obstruction of justice for firing Comey.  Hell, the hystericals wanted to pass legislation protecting Mueller’s job from a non-existent threat.  And then we come to our vaunted intelligence community – remember the Clapper report that got the Trump hysteria started?  And given all they knew that was not true, recall all the air time on lamestream media with guys like Brennan and Clapper giving regular interviews convincing people that they knew that Trump was “treasonous” and served as a “Russian agent” – in short, for the light thinkers in the land, the intelligence mavens abetted the TDS hysteria for the last 22 months.  Anybody going to hold these pernicious pilgrims to account?

[24mar19 update]  The Federalist reports ‘9 Years Into Common Core, Test Scores Are Down, Indoctrination Up’.  Of course, you would not hear about any of this from lamestream outlets.  Common Core was another one of President Obama’s big consequential lies to the American public – CC would “not only make America’s entire education system the envy of the world, but we will launch a Race to the Top that will prepare every child, everywhere in America, for the challenges of the 21st century.”  Well no, the result has been exactly the opposite.  But what has increased is the load of socialist crap that teachers deliver daily in the country’s union-dominated public schools.  And as reported before, teachers' continuing education under CC, from outfits like UnboundEd and EngageNY, trains them to bring leftwing brainwashing into every subject imaginable.  “UnboundEd’s training in reading and math instruction is ‘grounded in conversations about the roles that race, bias and prejudice play in our schools and classrooms.’ Its Standards Institute prepares educators to be ‘Equity Change-Agents.’ To become one, participants are told, they must first acknowledge that ‘we are part of a systematically racist system of education.’”  (H/T to reader, and more here)

Big corporations are good, so argues Canadian-American economist Rob Atkinson, president of Information Technology and Innovation Foundation, one of the world’s top think tanks for science and technology.  Atkinson points out some very surprising data about big vs small businesses not known to many.  Perhaps most important are their pay and benefit stats – big company workers make 54% more than workers for small companies, and both high school grads and those with some college make significantly more at big companies.  It all has to do with productivity and economies of scale which benefit not only the companies’ employees, but even more its customers and the public at large.  Atkinson argues persuasively that it has been technological advances that made big companies possible, and allows them to make enough money to keep innovating and pushing the technology peanut ahead.  Readers know my take on corporatism which appears to be the immutable destination of the big getting bigger to the biggest.  Atkinson is an advocate of big being prudently regulated by mindful regulations (he gives examples) instead of succumbing to the Left’s howl that big corporations should automatically be busted up by the government (e.g. a la Lizzy Warren who, nevertheless, has nothing against corporatism that contributes to her re-elections).  Anyway, before shooting from the hip on this, take a look at the data Atkinson presents (here) and in his 2018 book Big Is Beautiful: Debunking the Myth of Small Business.

The US advanced its Economic Freedom Ranking from 18th to 12th in 2018 (more here).  Now what puzzles me is whether this was due to Trump's tax reductions, regulation rollbacks, and other business-friendly policies and rulings; or was it really due to all the years of Obama's economic policies at less than 2% pent up annual growth that led to this sudden jump which just happened to coincide with Trump's presidency – it's really hard to tell. 

AG Barr's letter to Congress summarizing the Mueller report can be downloaded here.

[25mar19 update]  Kudos to George Boardman for his excellent column (here) highlighting Gov Newsom’s elitist overriding the will of the people on California’s death penalty, and also for pointing out a serious attempt to breach the First Amendment and the sanctity of our Fourth Estate in Congress’s assault on Fox News.  Our intrepid Union columnist and RR reader doesn’t make it easy for me in trying to paint him as progressive when now and then he comes out with a piece like this, which any good conservative commentator could have written.  Nevertheless, thank you Mr Boardman.