|
Yes, fantastic, in how much time compared to those algorithms?
In many cases speed and a moderate size reduction is a lot better than smaller but slower (and maybe by a lot, the paper doesn't say and I highly doubt AI is nearly as fast)
|
|
|
|
|
Gerry Schmitz wrote: (operations research). It's just more accessible without having to understand it.
Amen.
The three most dangerous things in industry are:
- A software engineer with an item of hardware
- A hardware engineer with a program
- A manager with a model
Freedom is the freedom to say that two plus two make four. If that is granted, all else follows.
-- 6079 Smith W.
|
|
|
|
|
Economies are far too complex to be modelled. Many economic models already exist, and none have good predictive value. Economies have far too many variables and no constants. The entire exercise is a reflection of "science envy" on the part of economists who think they can construct models in the same way as physicists. It is a total waste of time, made even worse by the fact that models can reflect deranged beliefs held by the model makers.
|
|
|
|
|
the far too many variables is precisely why i assume ai can be utilized .
|
|
|
|
|
The variables are largely based on human behavior, which is often unpredictable.
|
|
|
|
|
the unpredictability of human behavior is precisely why i believe ai can be utilized . isn't that what so called deep learning models do id est find patterns in vast data impossible for humans to process .
|
|
|
|
|
I suggest rather than starting at the Central Bank, let's pick a small town out of the way where the damage won't be catastrophic
I suppose you could run the AI in parallel to compare to existing controls.... might be interesting.
The thing is, hell has roads paved to it with good intentions. Going back to the early 80s, Reagan did "tax reform" and radically altered agricultural financing and destroyed the Savings and Loan industry, etc. The problem is that the government has far too much influence in the economy to make sense. Now you suggest a benevolent AI developed by government to make things better?
Charlie Gilley
“They who can give up essential liberty to obtain a little temporary safety deserve neither liberty nor safety.” BF, 1759
Has never been more appropriate.
|
|
|
|
|
yes . ai can be trained on vast data and so possibly foresee the unforeseen consequences .
|
|
|
|
|
"possibly foresee...."
fair enough. Let's start small. Other than an over whelming amount of hype, doom and gloom, terminator, the jury is still so far out that I suggest very small steps rather than pushing the red button (with a sign over it "do not push") to see what happens.
A very long time ago, some government gave a contract to Georgia Tech to come up with a super duper algorithm to help with traffic flow in Atlanta. Atlanta suffers from combined interstates at the center of the city (stupid right there) and multiple large venues. So, the algorithm was developed to manage traffic flow by automatically controlling traffic lights to "efficiently manage egressing traffic." It was an unmitigated traffic Armageddon disaster. Turns out that the side streets full of cars got tired I've waiting and said, "elephant it..." and ran the red lights.
start small, iteratively develop it, keep people who have to pay the bills to control the process....
Charlie Gilley
“They who can give up essential liberty to obtain a little temporary safety deserve neither liberty nor safety.” BF, 1759
Has never been more appropriate.
|
|
|
|
|
i am not surprised a human algorithm of a complex/incompletely understood phenomena developed in a previous era went haywire . i am surprised Phds failed so badly . in this era we now know about unforeseen consequences . to quote Donald Rumsfeld “There are known knowns, things we know that we know; and there are known unknowns, things that we know we don't know. But there are also unknown unknowns, things we do not know we don't know.” in a previous era such a thing would not have been spoken .
|
|
|
|
|
It hits the same tipping point that ML (AI "content generation") stuff hits.
Once you have it creating most of what it is learning it isn't really creating or learning anymore.
|
|
|
|
|
Unless that modelling correctly understands human behavior it will fail, just like all other economic central planners have failed.
|
|
|
|
|
perhaps ai can deduce human behavior just as it deduces other matters from vast data sets exempli gratia ai is beginning to deduce the rules of physic and obviously human conversation/chat etc .
furthermore i have some regard for social scientists whom investigate . it is unfortunate they are not consulted by governments whilst forming policy .
|
|
|
|
|
I think that you are much too optimistic.
It is not sufficient to collect a large mass of data and throw an "AI" at it. A model with a set of degrees of freedom (parameters) must be developed, and then the AI may be "trained" to adjust the parameters so as best to fit the data. This assumes that the parameters are not time-dependent, which is an even more difficult problem.
On the micro level, any national economy has a large number of degrees of freedom (underestimated as 1 degree per person). No model, AI or otherwise, comes even close to handling such a large number of variables. Furthermore, the "boundary conditions", i.e. interactions outside the national economy, are at least as complex. It may be possible that the smallest unit that may be modelled is the planetary economy.
On the macro level, economists have yet to identify the variables that are important to the economy, the proof being their inability to forecast anything but the most general trends.
EDIT: grammar
Freedom is the freedom to say that two plus two make four. If that is granted, all else follows.
-- 6079 Smith W.
modified 2-Oct-23 1:24am.
|
|
|
|
|
It also assumes people "never change" ... I "change" based on issues and have never stated "I always vote that way".
"Before entering on an understanding, I have meditated for a long time, and have foreseen what might happen. It is not genius which reveals to me suddenly, secretly, what I have to say or to do in a circumstance unexpected by other people; it is reflection, it is meditation." - Napoleon I
|
|
|
|
|
Gerry Schmitz wrote: It also assumes people "never change"
I wrote: "This assumes that the parameters are not time-dependent, which is an even more difficult problem."
Freedom is the freedom to say that two plus two make four. If that is granted, all else follows.
-- 6079 Smith W.
|
|
|
|
|
I got news for y'all... bots have been used in the markets for years now. Just about every market is manipulated too. So, whether it's to benefit a few people or machines, either way you're screwed.
And yes, I know you beat the odds. Don't mean they're not already botted, etc. though.
Jeremy Falcon
|
|
|
|
|
BernardIE5317 wrote: i just now learned of ai economic forecasting/modelling
So AI exists by consuming existing materials.
So who is presuming that the human economists are doing a great job?
|
|
|
|
|
It will probably be like Metal Gear. War is the quickest way to make a profit. And the machine will quickly learn that.
Without the military industry complex there would be no Federal Reserve. Without the Federal Reserve there wouldn't be a military industry complex. Two evil organizations that need each other.
|
|
|
|
|
|
"I've seen it," said Marvin. "It's rubbish."
(I saw it from the outside about a year ago.)
|
|
|
|
|
But do you have a terrible pain in all the diodes down your left side?
"I have no idea what I did, but I'm taking full credit for it." - ThisOldTony
"Common sense is so rare these days, it should be classified as a super power" - Random T-shirt
AntiTwitter: @DalekDave is now a follower!
|
|
|
|
|
Not all of them; but it's the replacement leg that's a bigger issue.
|
|
|
|
|
That's the NHS for you; even a robot won't live long enough to get treatment. 
Freedom is the freedom to say that two plus two make four. If that is granted, all else follows.
-- 6079 Smith W.
|
|
|
|
|
Indeed, the first million years are always there worst.
|
|
|
|
General 
News 
Suggestion 
Question 
Bug 
Answer 
Joke 
Praise 
Rant 
Admin 
Submit an article or tip
