patrick.net

Rin says

Wasn't chess settled already, back with Deep Blue and Kasparov in the 90s.

All that was settled was that computers could beat the best chess players, sometimes. And that was using conventional human time controls, a huge advantage for the computer.

Rin says

It's the top dogs in corporations who set the trends and if/when tools like IBM Watson infiltrate the white collar space, watch out!

I see your point. But the value the top dogs add themselves is in large part bringing all these human skills together.

Once they are just sales people sitting on a layer of commoditized automated operations, they may find their businesses themselves are becoming commodities for their customers.

marcus says

THe best chess players against the best computer chess programs ? Draw at best, for the computers (usually). SO no, I don't think the best computers beat the best chess players consistently.

You're kidding right?
Kasparov got his ass kicked in the 90's.
Now almost 20 yrs later, it wouldn't even be close.

marcus says

Rin says

Wasn't chess settled already, back with Deep Blue and Kasparov in the 90s.

All that was settled was that computers could beat the best chess players, sometimes. But not that they usually will.

Ok, but a decade & half has past and from what I'd heard, Deep Fritz has been able to keep the disparity to 2:1, even against the likes of a Kramnik and on less hardware.

Heraclitusstudent says

I see your point. But the value the top dogs add themselves is in large part bringing all these human skills together.

Once they are just sales people sitting on a layer of commoditized automated operations, they may find their businesses themselves are becoming commodities for their customers.

Yes, and the purpose of this 'sales', call it polite society *Boiler Room* antics, is to grow and merge with a larger house, down the road, and then, I get a bronze parachute while the principal partners get golden ones.

Heraclitusstudent says

You're kidding right?

Kasparov got his ass kicked in the 90's.

I guess if a slight victory in the rematch, using human time controls is what you call getting his ass kicked.

Rin says

Ok, but a decade & half has past and from what I'd heard, Deep Fritz has been able to keep the disparity to 2:1, even against the likes of a Kramnik and on less hardware.

I don't know that much about it, but I'm thinking that with more time, it would usually be a draw, or the human might still have an edge. The extra time helps the human a lot whereas it does nothing for the computer.

marcus says

Likewise trading will be done by humans with computers as tools. It already is.

Depends upon the firm.

Some famous ones like Renaissance Tech, heavily uses automated trading systems. They have a huge highly technical staff in the applied sciences, who continually tweak models and re-optimize algos. They charge a lot in management fees, which we've used to our advantage.

Others, like the one I'm in, use algos for position management and risk assessment but the traders rely more on their intuition and experience in making the final decisions. This mixture seems to work for us.

marcus says

I guess if a slight victory in the rematch, using human time controls is what you call getting his ass kicked.

Ok it was still close at the time. But not now.

The larger point is that a playing chess is essentially a search in a well determined space, and this is the type of thing computers are good at. Humans are still better if the space is extremely large, like go, but eventually progress in software means humans are losing.

What about trading?

What are traders doing? Sitting at a desk, getting specific information, and making decisions on simple actions, buy, sell etc... Already a large amount of the trading is done with computers, still controlled by traders, so far...

As every business transaction in the world is increasingly captured in computers in real time, the inevitable consequence is that the role of humans in trading will diminish and the role of computer increase until this is as much of a utility as routing phone calls.

Heraclitusstudent says

As every business transaction in the world is increasingly captured in computers in real time, the inevitable consequence is that the role of humans in trading will diminish and the role of computer increase until this is as much of a utility as routing phone calls.

Why do you think I'm trying to get out, so quickly. This isn't a field to be in, long term.

Afterwards, I'll be attending medical school and do my own research.

Then the crash comes.

Edit: Rin beat me to the punch with the last post.

Heraclitusstudent says

As every business transaction in the world is increasingly captured in computers in real time, the inevitable consequence is that the role of humans in trading will diminish and the role of computer increase until this is as much of a utility as routing phone calls.

I just don't know. A lot of market behavior is related to human behavior. Programs can be programed to predict human behavior, but when the humans are removed, what do you have then ? IS it more predictable ? Less? I don't know.

I understand your opinion. I just don't know. I would say at about the same time that AI gets really good, computers may trade better than humans. But there are some interesting paradoxes and bewildering thought experiments this leads to if you ponder it.

Computational power is relatively cheap, and there will be a lot of money looking for good returns. Which computers will be the ones that can best predict what this new set of circumstances lead to ? OR will it be humans that see it first ? I don't know.

Will we be polling computers for their sentiment to try to predict market tops and bottoms ?

marcus says

A lot of market behavior is related to human behavior. Programs can be programed to predict human behavior, but when the humans are removed, what do you have then ? IS it more predictable ? Less? I don't know.

I would say at about the same time that AI gets really good, computers may trade better than humans. But there are some interesting paradoxes and bewildering thought experiments this leads to if you ponder it.

Wasn't this the basis for Isaac Asimov's "Foundation" series? That mathematically speaking, the behavior of the masses could be predicted and manipulated throughout history?

marcus says

I just don't know. A lot of market behavior is related to human behavior. Programs can be programed to predict human behavior, but when the humans are removed, what do you have then ? IS it more predictable ? Less? I don't know.

Actually I think so far computers are changing the way things are done, and where the jobs are, but they have not caused massive unemployment. We are still relying massively on human labor at every levels. Even for repetitive physical tasks low wage labor in poor countries may be cheaper than robots.

The problem is real AI. We may well see rapid progress in terms of what AI can do, and such a technology would change the world radically.
This will not be a good news for most of mankind.
And I think it is coming.

Heraclitusstudent says

The problem is real AI. We may well see rapid progress in terms of what AI can do, and such a technology would change the world radically.

This will not be a good news for most of mankind.

And I think it is coming.

Yes, it's coming and it spells the doom, for white collar workers everywhere.

A big problem these days is that firms aren't hiring domestic scientists, engineers, etc., to perform the automation. They've outsourced that aspect, hiring people on the H-1B visa instead. More foreigners have been imported over the past decade to take tech jobs, than have tech jobs been created.

The situation is now so bad that even top grads from top schools can submit their resumes to firms that apparently are doing this sort of work, not even to receive the basic courtesy of a response. Large swaths of extremely talented people are going un-noticed, their skills going into a state of atrophy. Meanwhile, the lower-end, the government workers, seem to be doing incredibly well. As are the service sector workers.

Mark77 says

The situation is now so bad that even top grads from top schools can submit their resumes to firms that apparently are doing this sort of work, not even to receive the basic courtesy of a response.

The thing is that the owners and execs have devalued labor. Realize this, in the past, HR was called personnel dept, not human resources. Even the overall label, given to employees, is insulting and demeaning, like they're furniture and not people.

Now, with the world trying to garner science & engineering degrees, it's made it even easier for execs, to move work around, to accomplish their goals which is a certain amount of output with lower overhead.

The thing is that as long as we value business-y types, we will see never see a resurgence in American tech careers. Ppl like myself, who'd started in tech R&D, with a degree in applied chemistry/chemical engineering, will leave for B-school (a/o finance), patent law, or health care/medicine.

In fact, even within technology careers, ppl would rather become Oracle DBAs and system admins, managing data centers, than in programming and core R&D. And part of that is that it keeps 'em closer to management than tech, the idea being that R&D is the first thing which gets outsourced. A VP would still like his pet DBA, producing reports for him, and may not be comfortable, sending that job over to Asia, as he'd use that data for his personal career path.

"the development of information technologies and the automation of work"

that and the invention of intermodal cargo and global communication, both of which facilitating trade with cheap-labor / weak currency states like China, India, E Europe, and the Philippines.

http://research.stlouisfed.org/fred2/graph/?g=Bm8

blue is working-age population, red is workers in mfg and information (right axis).

We've lost 6M -- 30% -- of our jobs in just these two sectors since 2000. This goes beyond technology and has more to do with trade policy.

NAFTA, MFN with China -- we've seen TRILLIONS being sucked out of the paycheck economy thanks to free trade (with trade deficits).

http://research.stlouisfed.org/fred2/series/NETEXP

Heraclitusstudent says

I think so far computers are changing the way things are done, and where the jobs are, but they have not caused massive unemployment.

http://research.stlouisfed.org/fred2/series/USINFO

Bellingham Bill says

This goes beyond technology and has more to do with trade policy.

The problem is that this is what execs want.

Think about it, all the clamoring for the govt to increase STEM workers is just a smoke screen to offshore more STEM work.

In the end, executives do not like headcount. In fact, one VP I remember, said a bold face lie, "We like headcount! We want to grow!". And then, a year later, half that company was let go and the non-sales, non-consulting roles were sent to Asia.

Rin says

In fact, even within technology careers, ppl would rather become Oracle DBAs and system admins, managing data centers, than in programming and core R&D.

I thought that was because being the DBA pays better.

One thing to consider is that China is teetering on implosion. This is a natural consequence of decades of mercantilism. All of that manufacturing is going to shift to other countries.

zzyzzx says

Rin says

In fact, even within technology careers, ppl would rather become Oracle DBAs and system admins, managing data centers, than in programming and core R&D.

I thought that was because being the DBA pays better.

Why do you think it pays better? It's the closest one can get to management, a.k.a. data center management, while still retaining a technical role. One doesn't need to go back to school to become a DBA. Afterwards, if one attends enough meetings with execs, they become a database manager. The shift in pay had occurred, during the mid-90s, as a result of this phenomena.

indigenous says

All of that manufacturing is going to shift to other countries.

Yeah, but that's it ... other countries, like Vietnam, Malaysia, Philippines, etc. The jobs seldom come back to America.

Rin says

Yeah, but that's it ... other countries, like Vietnam, Malaysia, Philippines, etc. The jobs seldom come back to America.

The fact is that the largest manufacturing country in the world by far is the US, there is a reason for that, we are good at it.

When a shopper goes shopping for turbines, heavy construction equipment, airplanes that will fly 100s of people, with six sigma performance, they look at value for their money.

The value is measured in unit cost. If the product is not six sigma it does not get considered. Vietnam, Malaysia, Philippines, etc. have a hard time making stuff to six sigma standards. Let alone at a lower unit cost.

That being said those trinkets that you see in Walmart are stuff the US was not competitive in because of labor cost. Thanks to Uncle Ben the dollar is worth less than it used to be. That means that some of that stuff might be made domestically.

Before you turn you nose up at this understand that manufacturing in the US is more and more automated. When the price of a robot is at 100k the payment is around 2k a month at which point you say what sort of worker will I get for 500 a week.

The economy grows in unexpected ways if the government doesn't get in the way. Which it has been big time for the last couple of decades thanks to Bush and Obama. We desperately need another Coolidge.

indigenous says

Before you turn you nose up at this understand that manufacturing in the US is more and more automated. When the price of a robot is at 100k the payment is around 2k a month at which point you say what sort of worker will I get for 500 a week.

Yes, there's both automation and lower domestic shipping costs. Net result, fewer workers and the profits role up to the executive suites.

indigenous says

Vietnam, Malaysia, Philippines, etc. have a hard time making stuff

Yes, they don't need to manufacture goods, not used in Asia itself by multinational corps, but they can maintain a global code base and probably with a higher level of efficiency than the body shops in South India.

In fact, India's reliance on the body shop mentality, instead of expert IT solutions, has made itself irrelevant, for the next IT wave. I think that India will go back to its 1980 level, when global corporate America finds alternatives in Southeast Asia.

They were a paper Bengal tiger.

Ted Kaczynski was right about technology. It isnt just a collection of small gadgets anymore, it is a holistic system when your oven needs to connect to the WiFi. Technology can never become less efficient or useful, which will eventually leave human labor obsolete.

Rin says

Yes, there's both automation and lower domestic shipping costs. Net result, fewer workers and the profits role up to the executive suites.

You act as if companies don't fail as well? How many corporations have been around for over 100 years?

Rin says

Yes, they don't need to manufacture goods, not used in Asia itself by multinational corps, but they can maintain a global code base and probably with a higher level of efficiency than the body shops in South India.

What?

Rin says

In fact, India's reliance on the body shop mentality, instead of expert IT solutions, has made itself irrelevant, for the next IT wave. I think that India will go back to its 1980 level, when global corporate America finds alternatives in Southeast Asia.

What?

India's biggest problem is there corrupt bureaucratic government, that and a piss poor job of training people.

epitaph says

which will eventually leave human labor obsolete.

Poppy Cock

indigenous says

Rin says

In fact, India's reliance on the body shop mentality, instead of expert IT solutions, has made itself irrelevant, for the next IT wave. I think that India will go back to its 1980 level, when global corporate America finds alternatives in Southeast Asia.

What?

India's biggest problem is there corrupt bureaucratic government, that and a piss poor job of training people

Regardless of cause/effect, corporations will slowly stop the IT offshoring to India for better destinations.

Rin says

Regardless of cause/effect, corporations will slowly stop the IT offshoring to India for better destinations.

I thought Libertarians were negative but you mutts take the cake.

indigenous says

You act as if companies don't fail as well? How many corporations have been around for over 100 years?

When a company fails, its assets get absorbed by private equity, spun out, and merged some some other company.

As for 100 year firms, we have Pratt, GE, DuPont, Westinghouse, State St, etc.

Rin says

As for 100 year firms, we have Pratt, GE, DuPont, Westinghouse, State St, etc.

what percentage is that?

yes the bankruptcy process, so what?

indigenous says

Rin says

As for 100 year firms, we have Pratt, GE, DuPont, Westinghouse, State St, etc.

what percentage is that?

yes the bankruptcy process, so what?

You asked if there were century old companies. Someone even put together a list ...

http://en.wikipedia.org/wiki/List_of_oldest_companies_in_the_United_States

As for percentages, who cares, when today's JP Morgan Chase was a merger of Manufacturer's Hanover, Chemical, Chase, Bank One, JPM, Bear, Wash Mutual, and so on. That's what, an 8 to 1 adsorption?

Rin says

Manufacturer's Hanover, Chemical

BTW, I used to know mid-level ppl at the above firms, prior to the merger.

Well, all of 'em were whacked but I'd bet that the redundant VPs above them were given nice exit packages.

Rin says

As for percentages, who cares

The point is that there is a very high mortality rate for Corporations.

You have drank so much kool aid that you decisions are so full of hate that they are irrational.

I have to be careful of that with government, very few are more negative than the libertarians. It will color your thinking if you are not careful.

Rin says

Then, something changed ... the development of information technologies and the automation of work.

You need to reread history and see everythings changes constantly.

Your whole thesis is proven wrong.

https://www.youtube.com/watch?v=-B9k6FXCPOc

Rin says

Millennials will be the 1st generation to see the end of work

Perhaps more like the first stupid generations because they keep
ignoring the past... all the accomplishments...

little snotty arrogant pricks more like it...

Rin says

hen a company fails, its assets get absorbed by private equity, spun out, and merged some some other company.

As for 100 year firms, we have Pratt, GE, DuPont, Westinghouse, State St, etc.

many just die out... with no outside interest, assets sold to public.
always interesting to go to SV liquidation sales/auctions.

indigenous says

The value is measured in unit cost. If the product is not six sigma it does not get considered. Vietnam, Malaysia, Philippines, etc. have a hard time making stuff to six sigma standards. Let alone at a lower unit cost.

very true and people keep missing this. In short no matter how cheap the labor costs, the spoilage is too high which gets factored into good units which
does not lower unit costs, but increases.