Strange Times
Over at CBS they had a piece reporting that the price of gold has had a big drop. It is now selling for just over $1,500/ounce while it was over $1,600 during the Yukon Quest and Iditarod [several of the prizes involve ounces of gold].
The piece also talks about the drop in the price of oil, and the record highs of the stock market, while the yield on 10-year T-bills is down to 1.72%.
Gold usually drops when the economy is expected to improve, and the stock market matches that sentiment, but oil goes down when demand is expected to drop because of weak economic conditions.
The only thing that makes any real sense is the T-bills, which is most probably a result of people fleeing the Euro after the ‘Cyprus solution’ went into effect. T-bills would seem to be one of the few things that is selling well. If we had people in charge who knew what they were doing, we would be selling more T-bills, not destroying our economy to reduce the number that are available.
—————-
All Things Considered had a segment on data collection and mining today. They included a quote attributed to Albert Einstein: “Not everything that counts can be counted, and not everything that can be counted counts.” This incessant collection of data is pointless.
It was pointed out that the data doesn’t really provide any context, and search algorithms can’t determine them from the raw data. Google gets on my nerves when I decide to price something, mostly for other people because I can create better search terms to find what they are looking for, than they can. Every site that uses ads tied to the Google system will be filled to overflowing with banners and boxes for whatever I searched for. Amazon gets really annoying after I get something for my Mother. There is no way of really getting the point across that I really don’t want to buy any more CDs from artists that my Mother likes. I really miss the local record store where I could get this stuff with cash and not be pestered to buy more of it.
They interviewed an individual praising the utility of electronic medical records. Nice concept that is totally worthless because, as I was reminded on Friday, the bloody ‘health care professionals’ don’t bother to read it.
6 comments
Ah yes. Data. I’m drowning in it. Do I really need to know the voltage of the 5 volt power supply in the video recorder every 5 minutes for the last 12 months? What in the world am I going to do with that data? What questions am I going to answer with it? I did a quick prune and knocked 55% of the data out of the database without even trying — just deleting duplicates and replacing them with start-stop timestamps (with a slight bit of hysteresis to deal with normal ADC jitter for determining what constituted a duplicate). My boss plaintively asks, “do we really need all this data?” Well, no. Just knocking blatant duplicates out is only the start. I need to go through it and see what we do need before pruning further. But I know for a fact that I don’t need 100,000 records showing that my 5 volt power supply has, in fact, been providing 5 volts of power at every 5 minute interval for the past year :twisted:.
You can’t pan for gold at the mouth of the Mississippi River… all you get is mud. Just sayin’.
This was my major complaint about the ‘Total Information Awareness’ project. We had all of the data we needed to discover the 9/11 attack, but we lacked the analysts to find and report it. Gathering more data is a brain dead idea until you have the ability to analyze what you already have.
Clinton was increasing counter-intelligence resources, but the Shrubbery put a stop to it when he came into office. Richard Clarke told the sordid story to the 9/11 Commission and the public, but the Hedgemony insisted it needed to vacuum up more data.
I think that I might, at most, include a single-byte flag field for things like that to cover any exceptions, but I certainly wouldn’t have a floating point field for the actual voltage.
Of course I did a lot of batch things which make that a lot easier than doing real-time data acquisition. I did some work for a researcher who had an instrument that monitored a number of different values from an experiment 24/7. He only really wanted exception reporting of all of the values. Basically, if something changed he got values from 5 minutes before the change until 5 minutes after things went back to normal. It worked out to a report about every six weeks, and he didn’t have to waste time looking at reams of worthless data. Given the cost of disk space back when I did that, it helped his grant money stretch a lot longer than it would have.
The quote is probably not of Einstein, though he may have been aware of the maxim at some point: there’s a photo (doctored?) of Einstein with the quote written on his blackboard. For one rare time, snopes.com was not particularly helpful, but quoteinvestigator.com had a lot of info.
DHS will not collect, or at least will not release to the public, any data that tends to debunk the asserted need for its own existence. Their name acronym is too long: DHS could and should be shortened to BS.
Actually it is almost a word for word translation of KGB with Department substituting for Committee, and the briefs are quite similar with tentacles in all kinds of seemingly unrelated areas.
As for the quote, I used ‘attributed’ because the concept matches, but the wording doesn’t. The wording is too ‘native speaker’ for Einstein, IMHO. The NPR story calls it a quote, but that is a real stretch. It is listed by a number of sites, but I haven’t seen a source specified.
Bryan, the maxim for the Big Data people is “collect everything because you never know what will be useful.” Of course, this sells Big Data services so of course they say that :twisted:. My current task is to go through all the sometimes-half-baked stuff that the kiddies did, and turn it into a production system, part of which is to quit gathering so much #$%! *JUNK* data!
It sounds like the Abteilung der Staatssicherheit has the same philosophy as the kiddos — gather everything then apply Big Data techniques to mine it for pearls. But if you got a pile of cow flops, all you’re gonna get trying to mine that for pearls is one hella stench attached to your clothes…
Big Data wants to show all of the PowerPoint slides that you can generate of all of the worthless data, and then the meaningless maps that you can draw with the same crap. The really is that all of that stuff simply reports static, not information.
Going with your mining analogy – there definitely is still some gold at the mouth of the Yukon River, but it would cost tens of thousands of dollars in time an equipment to extract an ounce out of the river, and that ounces was selling for $1400 [it dropped another $100/ounce today]. If you had a technique that would extract the ounce for $100, it would be worth doing, but no before.
If they had developed the tools and techniques that were capable of dealing with the old level of data collection, increasing might have made some sense, but when you can’t process what you already have, why bring more? These people believe in magic, because there is no logical justification for TIA.