This is a pretty hot topic today and I write about it a lot here, have even asked the Forbes guys which articles were algorithmically have been produced and where did the humans write. That’s out there and other sites besides Forbes are using it as well. The charts used here in the video showing GDP from the past and it is good for educational purposes. Productivity is at an all time high, why, because technology and machines are doing a lot of it. This creates a problem for standard metrics. I like when he ventures into how economists measure value and “getting things for free”. Zero price means zero waste. Music industry is half the size he states years ago. He states we have not reached the end of growth. Machines are doing it.
He makes one mistake though on the “digital” side that things can be replicated with perfect quality…ok that works for music but not date with algorithmic formulas that do a lot of things out there. Just look at the markets for errors, thousands of them, daily, mostly small but the big ones hit the news of course, but a small one may not hit the majority, but it could affect you. Lot’s of flawed data out there.
We do have ways to measure that we never had before and machines do it faster. He does a good job in briefly going over the linear world and how we are geared in that and the other side is the non linear world. Ideas don’t get used up and he says exactly what I say about software building on itself so I won’t bore you with a lot of that as it has worked that way for years.
He talks about the break throughs we are seeing by combining all of the above in the image. He talks about IBM Watson and machine learning and how it learned faster than any human. Remember though those who control the input, models and how Watson is used control the money, like Citibank is doing. Less than a year ago it went to work on Wall Street, to runs those models, query and work faster than humans with analytics. Again, remember Watson kicks out what it has been programmed to do and use. As you can see from my link title, Congress has not figured this one out yet and so they get out modeled right and left and duped on what is accurate and what has value, they are out tooled here. Here’s what happens when machine learning makes an error, Google’s machines said I was a real duck..in other words all the meanings, etc. were not “learned” if you will and it went with what it had…it thought I was a real bird.
IBM Watson Going to Work At Citigroup on Wall Street–Congress Didn’t See Big Data As A Tool (Hadoop Framework) When They Had Their Chance…For Consumers The Attack of the Killer Algorithms–Chapter 22
What is left out of this discussion is the disruption of data selling, guess what works here…machine learning…fewer people have jobs too. Some of this is due to technology and some of it is flawed data that misses a lot of our talented people behind. You can also watch a video at the link below that also describes big data being used out context as big companies don’t have a clue on where non linear value lies and banks make some of this up or just lie with models…LIBOR anyone.
Big Data/Analytics If Used Out of Context and Without True Values Stand To Be A Huge Discriminatory Practice Against Consumers–More Honest Data Scientists Needed to Formulate Accuracy/Value To Keep Algo Duping For Profit Out of the Game
Economists need to learn about “free” and the “data selling” epidemic that takes place to allow them some better accuracy and realism as the past offers history for education but not much value for gating the future.
Modeling for Inequality With Segmentation, Insurance Industry Uses Backwards Segmentation As Some Models Stand to Threaten Overall Democracy
Everyone is attacking Obamacare, and it’s not the overall program itself but rather the complexities of models created by insurers that is one big chunk of problems and you can even see that with the recent frustration with the Inspector General with wanting insurer CEOs to “certify” to CMS that their information is correct and accurate because it doesn’t show up that way now in all areas on Medicare. Gov.
Not ObamaCare That is Failing, It’s the Models and Subsequent Algorithms that Execute Within IT Infrastructures Intersecting, Changing And Conflicting– The Affordable “Complexities” Act…
All in all the video covers a lot of areas relative to machine learning as best as one can in the 20 minutes of a TED presentation. We still need to balance our intangible and tangible worlds. People are racing against the machines and some are losing. We could have some of the best manufacturing in the US if only the other verticals are given more technology instead of just selling data.
Think about this if we didn’t have the threat of always having a profit and using data against us, what we could do, but we don’t have that world sadly at all. We still blame shift and don’t work well as teams and use technology to segregate now in many areas, whether you are the evil fat person, the sedentary person or whatever, that is how we see a lot of technology used. We have banks and companies that profit and make billions off selling flawed data and we end up being their “free” labor force to fix it as we are stuck and can’t buy a car, get a job or much else until we spend out dime fixing corporate America’s profits where they don’t give a damn about the average US consumer anymore. Watch the video above at the Modeling for Inequality link above for more on that. License and tax data sellers to make them accountable? Yes the way to go and step one of moving some money back to the 99% side for all of us. BD
Time Has Come to License and Tax the Data Sellers of the Web, Companies, Banks, Social Networks..Any One Making a Profit-Latest Microsoft/Google Privacy War Helping the Cause –Consumers Deserve to Know What Is Being Sold and To Who in a Searchable Format