Guild of Project Controls: Compendium | Roles | Assessment | Certifications | Membership

Tips on using this forum..

(1) Explain your problem, don't simply post "This isn't working". What were you doing when you faced the problem? What have you tried to resolve - did you look for a solution using "Search" ? Has it happened just once or several times?

(2) It's also good to get feedback when a solution is found, return to the original post to explain how it was resolved so that more people can also use the results.

A Question For a Masters - Will Computers Take Ove

22 replies [Last post]
Darrell ODea
User offline. Last seen 15 years 41 weeks ago. Offline
Joined: 27 Feb 2002
Posts: 174
Groups: None
Folks,

Am completing my masters, doing the thesis bit at present.
Have a question to ask as part of my research, set out below. If any of ye are interested in answering, please forward your comments to me or, post directly to this site.

My email address is as follows: odead@sisk.ie.

Thanks in advance,
Darrell

“A fairly ridiculous statement, in some peoples eyes to make, would be, that computers, expert systems, knowledge based systems, or artificial intelligence, will to a large extent replace our key industry leaders, where policy and decision making are concerned. Specifically for the Construction Industry.

How strongly do you disagree / agree with the above statement?

 Strongly Disagree
 Mildly Disagree.
 Both Agree & Disagree.
 Mildly Agree.
 Strongly Agree.
 Unsure.

Please give a reason or reasons for your answer.


Replies

Daya Sugunasingha
User offline. Last seen 13 years 37 weeks ago. Offline
Joined: 27 Feb 2002
Posts: 99
Groups: None
Darrell
Quite well rounded up I think
Darrell ODea
User offline. Last seen 15 years 41 weeks ago. Offline
Joined: 27 Feb 2002
Posts: 174
Groups: None
Folks,

Thanks again for all yer interesting commentary.
Some very useful notes indeed, and from some very polished individuals indeed.

Cannot help but re-comment.

Indeed: Computers can now beat perhaps well over 99% of the "human" population as a whole at chess, one would have to admit?
Grandmasters account for perhaps a very small fraction of the human population, although might I suggest, any human given time and enough coaching and training could potentially reach grandmaster level. Thus increasing the opportunity or potential to beat the best of computer chess programmes.

On The Take Over of AI in our (Construction) industry:
Perhaps in the "developed" world, on closer examination we can see this happening?
One would have to admit that, factory production of building or construction materials, is driven physically, not by human toil any more, but by machines. On site erection of components, as time goes by, is mainly by mechanical and electronic devices? I have seen robotic breaking devices, albeit controlled at a distance by a human. As time goes by we no longer require the level of human labour on sites that we once depended on. Perhaps econimics played a big part in this, as opposed to the need for an easier life?

Perhaps because a year is a relatively long time in terms of the human life span? It is a lot more subtle to us than we can recognise.

Darrell
Philip Jonker
User offline. Last seen 12 years 51 weeks ago. Offline
Joined: 7 Nov 2004
Posts: 852
Groups: None
On Second thoughts....

The problem exists in the planning industry and the most people don’t understand the technology they are running on.

The technology is DBF database technology which is 25 years olf, therefore they do not understand the ease in input in terms of modern techonology, like Excel, which already is old.

It’s a question of learning and keeping up with the times and the past.
Stephen Devaux
User offline. Last seen 29 weeks 3 days ago. Offline
Joined: 23 Mar 2005
Posts: 655
Hi, Philip.

Sorry, but "Mildly Disagree" was one of the options offered by Darrell in the original post. That’s my selection, and I’m sticking to it to the death!

I guess it sort of depends on what timeframe one is looking at. I certainly don’t see computers "taking over" this function in this century, but beyond that, who knows.

As for the chess metaphor, when MIT’s McHack defeated Ben Landey, for the first ever victory by a computer over a human, Bobby Fischer looked over the game and pronounced it "a disgrace to the human race." (This was accurate -- I knew the late Ben Landey; very nice man, absolutely horrendous chess player! The antithesis of Fischer. . .) The progress of 40 years, from McHack over Landey to Deep Blue over Kasparov, may seem huge; but it is probably nowhere close to the quantum leap that would be needed to take computer chess to the next level (say, beating grandmasters in every game with White and never losing with Black.) In fact, if you played over the games of the Deep Blue/Kasparov match, you know that Kasparov made some blunders that even Ben Landey might have avoided! I seriously doubt that Deep Blue would have gotten a plus score in the recent Linares tournament, for example.

All of which is a digression. Someone once observed that programming computers to play chess is easy; it’s programming them to play table tennis that’s hard! The fact is, I don’t think computers will be able to make PM decisions on their own, only support the decision-making process much more than currently.
Philip Jonker
User offline. Last seen 12 years 51 weeks ago. Offline
Joined: 7 Nov 2004
Posts: 852
Groups: None
[double post deleted]
Philip Jonker
User offline. Last seen 12 years 51 weeks ago. Offline
Joined: 7 Nov 2004
Posts: 852
Groups: None
Hi Stephen,

Thanks for re-surrecting yhis thread. It is not possible for a computer to think for itself. It needs human input. What you say is quite correct. But you must have a clear stand point on the subject, and "midly disagree" is not a clear stand point.

The fact that Kasparov (World Chess champion) has been beaten by the IBM Blue super computer is a generally known fact. It doesn’t make the computer cleaverer that Kasparov, it just calculates the potential moves, which ranges in the billion worth op potential moves, which then takes the most suitable part towards winning.

You statement about the "speed of light". DEFINE IT!

Einstein worked on 3 times ten to the power of ten centimers per second, you’ve got other supposiable brilliant people trying to calculate the same number and they come to figure like 2,999797979 for example. But they forgot to take the effects of the vaccumm into consideration.

How do we get to the boundary of human imagination and thought, was Einstein right? or the people that tried to find his mistakes/errors?

The question of AI is not a mild question so there is either a correct: They are taking over (computers) or NOT!

Let’s stop the symantics and speak honestly!

I like your posting!
Stephen Devaux
User offline. Last seen 29 weeks 3 days ago. Offline
Joined: 23 Mar 2005
Posts: 655
Please excuse my disinterring some old threads -- I’m new to the forum, and, frankly, am quite excited about many of the topics being discussed here. One does not find anything close to this level of sophistication in most PM discussion fora.

A couple of points:

Someone wrote: "The possible moves in a chess games if finite, so that if you have the ultimate computor, it can beat anyone at chess."

Even the ultimate computer will be, presumably, limited by the speed of light. Indeed, it is possible that the current approach to chess computer programming may have reached close to its optimum level, as it’s pretty much of a "grind-through-all-the-options" approach. With each ply-level deeper the computer looks, the number of options it must examine increases logarithmically. To go from looking two moves ahead to three moves ahead was relatively simple. But now the programs are routinely examining all options five moves ahead, sometimes more. Thus a huge increase in speed would be necessary to look even one ply deeper. The next advance is likely to require a better simulation of "human judgement."

This is relevant to the discussion because human judgement is the stumbling block for a project decision "expert system" solution. As a result, my answer to the original question would be "Mildly Disagree".

But, as I mentioned in the final chapter of a book I published five years ago, I think it would be relatively easy to improve on the current status of computer support, especially in the arena of resource targeting and leveling (especially on a multiproject basis). It would require, however, new input fields that reflect a change in project planning/management to recognize that the purpose of every project is investment, and that stress maximized value-above-cost (the great oxymoron: "project profit"!).

What I suggested in my book was a software package that quantified such value through user input, and that would crank for several hours, or over a weekend looking at all options and providing each with a "score". The user would then arrive on Monday morning, examine the results for the top 10 scoring solutions, and either select the one he/she likes best or move on to examine solutions 11-20, etc.

Such a system would not be an expert system, but it would provide much more computer support than anything currently out there.
Philip Jonker
User offline. Last seen 12 years 51 weeks ago. Offline
Joined: 7 Nov 2004
Posts: 852
Groups: None
Hi Darrel,
I wasn’t getting at you, and there is nothing like a simple mind. My computor is playing up today, thus, the four postings. How can one rely on a computor for decisions when they make mistakes like that.

Regards

Philip
Darrell ODea
User offline. Last seen 15 years 41 weeks ago. Offline
Joined: 27 Feb 2002
Posts: 174
Groups: None
Philip,

Thanks for that, I for one have a simple mind.
Philip Jonker
User offline. Last seen 12 years 51 weeks ago. Offline
Joined: 7 Nov 2004
Posts: 852
Groups: None
Hi Guys,
A couple of points.
Firstly on traffic lights, they were there long before computors and were run bz simple timer switches, so the point of them having any intelligence is invalid. Computors have just made their operation a bit more sophicated and their reliability less.
Darrell’s point on chess does not take into account the difference between computing power and speed from 1990 to 2002. I recently read an article about the Cray computor being the first computor to beat one of the grandmasters. The possible moves in a chess games if finite, so that if you have the ultimate computor, it can beat anyone at chess.
Coming back to AI I think it has its uses, but more as a tool in decision making than as the actual decision maker. One of the problems is the amount of data required to make certain decisions, one must never forget the complexity of the human mind. Secondly is keeping all the data up to date, eg something could be happening just before the decison is made, that will that could invalidate the decision, because the data has not been updated. Remember humans have senses ie hearing, sight and feel, that assists us in making decisions.
Take marketing, one of the major tools in business, try putting it at the entrance of a shopping mall to stop shoppers and ask them questions.
It still remains an interesting topic.
David Bordoli
User offline. Last seen 5 years 31 weeks ago. Offline
Joined: 8 Apr 2002
Posts: 416
My original comment of ’never say never’ may have appeared flippant but history shows us that learned minds of the past have often said ’never’ or ’impossible’ and have ultimately been proved wrong.

Such developments may not be in our comprehension or in our lifetime (or at all) but to dismiss as ’never’ I think is rather dangerous. Maybe its akin to being able to ’not think out of the box’, that is, to be constrained by a set of beliefs that reflect only current knowledge and experience.

Whilst I have a little difficulty in using the word ’learned’ in the same reply as mentioning HRH Price Charles he certainly has his doubts, or fears, about the progress of grey sludge nanotechnology. I haven’t followed that argument but my natural instinct is to put it in the same box as his views on modern architecture (well, my biased views on what his view might be!)

Regards

David
Darrell ODea
User offline. Last seen 15 years 41 weeks ago. Offline
Joined: 27 Feb 2002
Posts: 174
Groups: None
Thank you Bernard.

Your beliefs are reflected and agreed upon by a majority of people, I believe.

I do not hold or am trying not to hold a specific belief, just trying to establish facts, with a potential outcome.

You say & these are your words:- "humans must engage in creative/proactive thought/action across an unbounded/unlimited set of circumstances", when did you or me or anyone that we know do such a thing. 99.99% of what we do and think every waking moment, is someting we have done sometime before, it is programmed into us. Nature or nuture. There is very little creativity or productive thought or action(new) taking place. And we are all "very limited & very bounded" by our "Programming".

It is irellevant, that decision making is rational or irrational, right or wrong and so on.

Intuition is "the power to see truth directly without reasoning".
But the promble with that is humans instinctively reason, although we sometimes do it so swiftly, it seems like we are not thinking.
So by its definition the only thing that can do this is AI. Now there is a thing.
Bernard Ertl
User offline. Last seen 6 years 50 weeks ago. Offline
Joined: 20 Nov 2002
Posts: 757
While I understand the point you are driving at, I believe that you are limiting the viewpoint to support your argument.

Human decisions can be rational or irrational. Rational decisions can be intuitive or logical. Both methods have validity and I do not believe that AI will ever be able to model intuitive decision making. Do you believe that AI will be able to judge the potential for irrationality amongst stakeholders in a situation?

When you consider that humans must engage in creative/proactive thought/action across an unbounded/unlimited set of circumstances, I cannot buy the arguement that AI will ever be able to replace a human as a business manager. That is, unless you are suggesting that AI will one day be sophisticated enough to completely and accurately model the human brain which we still do not even fully understand.

Do you believe that AI will be able to establish personal relationships with people that will allow it to conduct business in a ’good old boy network’?

Bernard Ertl
InterPlan Systems - eTaskMaker Project Planning Software
Darrell ODea
User offline. Last seen 15 years 41 weeks ago. Offline
Joined: 27 Feb 2002
Posts: 174
Groups: None
Bernard,

Yes, many computer programmes are bounded by a set of limited possibilities or outcomes.

Taking your computer chess anology into consideration, as a chess player myself, can conquer a Kasparof computer game, that was bought for me back in 1990. The most recent computer game that I purchased for my PC back in 2002, I have yet to beat at any level whatsoever. Thus to date I have not have had enough programming, be it simple or complex, to enable me to achieve my desired outcome.

My own experience in itself although potentially limitless, at present is very finite, like many computer programmes. I have the capacity to expand, stretch & learn, but I only know what I know, through and up to my current programming.
Be it a combination of both Nature(Genetically) or Nurture(Conditioning). It is all still programming.

A decision is a decision is a decision, be it complex or simple. Wether it be right or wrong, correct or incorrect is not relevant. As per the Human race, we make decisions and accept decision and instruction every day and at every turn. Some are made by us, some are made by others and some are made and given by AI, be it at any level.

Some AI systems are very complex indeed and some very simple, like people.

All decisions and by the nature of their being an actual decision to make, or to deduce someting down to being able to make a decision,are very simple to make. Even in business decision making is finite, and very bounded. If it was infinite and unbounded (ie. The individuals experiences, conditioning and programming), then nothing would be done.

Decision making is only decision making and nothing more.
Although the actions that follow are profound, sometimes anyway. It doesnt matter whom or what made them, just that they were made.

Respect,
Darrell
Bernard Ertl
User offline. Last seen 6 years 50 weeks ago. Offline
Joined: 20 Nov 2002
Posts: 757
There are several systems (programs) for controlling traffic lights. Some are simple programs that switch colors on a timed pattern. Others are more complex and have hueristics that take input from pressure plates to identify which sections of an intersection have traffic waiting. In all cases, however, the lights are controlled by a fairly simple computer program.

The system is not designed to take all inputs and evaluate scenarios encompassing subjective variables beyond its input. I’m not an AI expert, but I have a hard time envisioning how an AI system will be able to ’think outside the box’. Much like the best chess programs, there are limits to what software can do even if it can process millions of possibilities faster than a human.

So while I agree that computers can aid in simple decision making, your original premise was a much stronger proposition and there is no logical reason to extrapolate from the simple to the complex IMO. The main problem is one of bounded versus unbounded problems. In chess or traffic signalling, the range of decisions is finite or bounded. In a business, the range of decision making is an unbounded situation.

Bernard Ertl
InterPlan Systems - eTaskMaker Project Planning Software
Darrell ODea
User offline. Last seen 15 years 41 weeks ago. Offline
Joined: 27 Feb 2002
Posts: 174
Groups: None
Thanks Guy.
And thanks to all of you who spared the time to answer.

Another question perhaps.
Do traffic lights make decisions (or the software behind them)?
If a light is red, and you "decide" to stop, as intimated by the colour showing, are you following an instruction, based upon a decision made by computer software (be it a pre-programme designed by a human) programme?

Is there a decision made here?
Is it by a "Low" level of Artifical Intelligence?
And do we follow these instructions & decisions?

Answers on a post card.....

Guy Hindley
User offline. Last seen 2 years 36 weeks ago. Offline
Joined: 28 Sep 2001
Posts: 91
Groups: None
I would agree with Bernard - Never. I will however concede that the role of computers, AI, et al will increase, but they need human brains behind to create the knowledge, and certainly verify data is correct and make the decisions. However good these applications become a human brain has to make the decision as to whether or not accept the decision the computer is suggesting is made/ taken. There is certainly a place for such systems to aid Planning and the management of those plans and schedules
David Bordoli
User offline. Last seen 5 years 31 weeks ago. Offline
Joined: 8 Apr 2002
Posts: 416
Never say never?

:-)
Bernard Ertl
User offline. Last seen 6 years 50 weeks ago. Offline
Joined: 20 Nov 2002
Posts: 757
AI will never be sophisticated enough to make decisions with political goals/ramifications "in mind". :)

I don’t see people entrusting computer systems to make business critical decisions for them. Computers and software are just tools.

Bernard Ertl
InterPlan Systems - eTaskMaker Project Planning Software
David Bordoli
User offline. Last seen 5 years 31 weeks ago. Offline
Joined: 8 Apr 2002
Posts: 416
Hi Darrell

Have you had a look at the work of Emil Laptali (not sure if I have spelt her first name correctly)? She did some stuff on programmes and AI sometime back (only know because I helped verify some of the stuff).

And 12 years ago I was getting excited about the work of Dr. Osama Moselhi (Concordia I think) and neural networks.

Trouble is I think I am a bit our of date and out of my depth!

Regards

David
david.bordoli@gvagrimley.co.uk

I am a bit out of my depth on the sort of thing you are talking about... policy making!
Darrell ODea
User offline. Last seen 15 years 41 weeks ago. Offline
Joined: 27 Feb 2002
Posts: 174
Groups: None
Just a comment to/on my own question.
Am to date, getting stong opinions about the possibility of any Artifical Intellegence (AI), being able to make decisions and formulating policy. Most people to date think, thus far that the possibility in itself is ridiculous. Thus if you have another view, I would love to hear it. Any views expressed will be taken into consideration of course. "Planners" and not to stereotype them too much, generally, think a little bit more outside the box, again only a view.

So go on, hit me with your best or worst.

Regards,
Darrell
Raj Maurya
User offline. Last seen 21 weeks 1 day ago. Offline
Joined: 17 Jul 2004
Posts: 132
Certainly Computers and sofwares are tool that can be utilised for presentations & calculations that makes easy to understand others who are not familiar with some one own’s creative mind. Critcal business decisions can be analysed through it.