SQLServerCentral / Editorials / SQLServerCentral.com / The Robot DBA / Latest Posts

RE: The Robot DBA

Jon Russell — Thu, 08 Sep 2011 17:03:51 GMT

[quote][b]SQLRNNR (9/7/2011)[/b][hr][quote][b]jcrawf02 (9/7/2011)[/b][hr]Ideally, using automation to replace workers would just require different and/or higher-level skills. The need for people to do *something* will never go away, the skills required will just change.[/quote]Agreed. Skillsets will change and there should be that type of change. Making better applications that can work faster and more accurate is good. In order to make those apps and to continue to support them, we need to continue to evolve our skillset too.[/quote]Yes, you both nailed it. This is actually an old economic topic. A specific task or even a job may be automated away, but that leaves an opportunity to invest in new innovation and efficency improvements by you or others.

RE: The Robot DBA

umailedit — Thu, 08 Sep 2011 00:32:49 GMT

Man I wish someone will automate my Jobs. Seems to me, I have enough work to last several lifetimes.

RE: The Robot DBA

David.Poole — Wed, 07 Sep 2011 16:33:41 GMT

I used to work for a company supporting other peoples software. I remember one of my colleagues looking at the original implementation of .NET and viciously asserting that "this will kill off all those cowboy developers".Similarly I can remember a comment on this site about clustering SQL Server. In SQL6.5 to SQL2000 it went from a "hey wow" skill to "So what"!Read Daniel Pink's Drive. He talks about human motivation and why carrot and stick is only really useful for dull mundane jobs that should be automated. People are self-motivated and enthusiastic when their creative energy is engaged.Most DBAs I know spend a considerable amount of time figuring out how to automate stuff so they can concentrate on the interesting stuff.SQL Server makes a wide range of DB tasks very simple. The problem comes in that people assume that something that is simple in SQL Server means that other niche products are simply bigger, faster versions of SQL Server requiring no particular skill.

RE: The Robot DBA

Steve Jones - SSC Editor — Wed, 07 Sep 2011 11:46:01 GMT

[quote][b]mtillman-921105 (9/7/2011)[/b][hr]Heh, I'm with you one this one Steve. We can't even automate loading a random flat file into a system. Over a decade ago, I remember hoping that one day such tasks would be unnecessary. [i]Wrong![/i] :-P[/quote]Not without a DBA around 30% of the time when things go wrong :-P

RE: The Robot DBA

cfradenburg — Wed, 07 Sep 2011 11:43:14 GMT

[quote][b]Revenant (9/7/2011)[/b][hr][quote][b]cfradenburg (9/7/2011)[/b][hr][quote][b]majorbloodnock (9/7/2011)[/b][hr]What is less believable is the automation of contract negotiation; the question of "do I want to do business with this company?" can hinge on all sorts of intangible and subjective criteria which are unmeasurable even if a person can instinctively weight them up in the mix.[/quote]But what percentage of the time is the wrong decision made based on those subjective criteria?[/quote]About forty percent. Several comprehensive (and expensive) studies agree that if you cannot reduce a decision to a few simple numbers on a spreadsheet and you weigh in subjective criteria, managers get about 60 percent decisions right, no matter how many consulting bucks and effort go into that decision.That's 10 percent better than to flip a coin.[/quote]That's greatly reassuring to me. Both the fact that most decisions are right and the fact that pouring money into consultants doesn't help much. I still wonder how many decisions have the "wrong" decision made for a subjective reason when it could be reduced to numbers. I'm not saying that subjective reasoning doesn't still have it's place. After all, it's worth taking a risk on someone because you want them to succeed even though the numbers don't bear it out. Or not working with someone because you know they'll be a pain.

RE: The Robot DBA

SQLRNNR — Wed, 07 Sep 2011 11:33:16 GMT

[quote][b]jcrawf02 (9/7/2011)[/b][hr]Ideally, using automation to replace workers would just require different and/or higher-level skills. The need for people to do *something* will never go away, the skills required will just change.[/quote]Agreed. Skillsets will change and there should be that type of change. Making better applications that can work faster and more accurate is good. In order to make those apps and to continue to support them, we need to continue to evolve our skillset too.

RE: The Robot DBA

Chris Harshman — Wed, 07 Sep 2011 10:41:12 GMT

[quote][b]Revenant (9/7/2011)[/b][hr]...managers get about 60 percent decisions right, no matter how many consulting bucks and effort go into that decision.That's 10 percent better than to flip a coin.[/quote]heh-heh... that makes me think more of Isaac Asimov's "The Machine that Won the War", which I think is closer to how things really go in the business world. ;-)

RE: The Robot DBA

majorbloodnock — Wed, 07 Sep 2011 09:22:22 GMT

[quote][b]Revenant (9/7/2011)[/b][hr][quote][b]cfradenburg (9/7/2011)[/b][hr][quote][b]majorbloodnock (9/7/2011)[/b][hrWhat is less believable is the automation of contract negotiation; the question of "do I want to do business with this company?" can hinge on all sorts of intangible and subjective criteria which are unmeasurable even if a person can instinctively weight them up in the mix.[/quote]But what percentage of the time is the wrong decision made based on those subjective criteria?[/quote]About forty percent. Several comprehensive (and expensive) studies agree that if you cannot reduce a decision to a few simple numbers on a spreadsheet and you weigh in subjective criteria, managers get about 60 percent decisions right, no matter how many consulting bucks and effort go into that decision.That's 10 percent better than to flip a coin.[/quote]You surprise me. In fact, I'm somewhat comforted that, averaged out, we get significantly more subjective decisions right than wrong.Also, given how little I trust some people's judgement, that must mean there are quite a few people who're really pretty good at subjective decision-making to bring the average back up again. ;-)

RE: The Robot DBA

mtillman-921105 — Wed, 07 Sep 2011 08:57:59 GMT

Heh, I'm with you one this one Steve. We can't even automate loading a random flat file into a system. Over a decade ago, I remember hoping that one day such tasks would be unnecessary. [i]Wrong![/i] :-P

RE: The Robot DBA

Revenant — Wed, 07 Sep 2011 08:51:54 GMT

[quote][b]cfradenburg (9/7/2011)[/b][hr][quote][b]majorbloodnock (9/7/2011)[/b][hrWhat is less believable is the automation of contract negotiation; the question of "do I want to do business with this company?" can hinge on all sorts of intangible and subjective criteria which are unmeasurable even if a person can instinctively weight them up in the mix.[/quote]But what percentage of the time is the wrong decision made based on those subjective criteria?[/quote]About forty percent. Several comprehensive (and expensive) studies agree that if you cannot reduce a decision to a few simple numbers on a spreadsheet and you weigh in subjective criteria, managers get about 60 percent decisions right, no matter how many consulting bucks and effort go into that decision.That's 10 percent better than to flip a coin.

RE: The Robot DBA

cfradenburg — Wed, 07 Sep 2011 08:40:41 GMT

[quote][b]majorbloodnock (9/7/2011)[/b][hrWhat is less believable is the automation of contract negotiation; the question of "do I want to do business with this company?" can hinge on all sorts of intangible and subjective criteria which are unmeasurable even if a person can instinctively weight them up in the mix.[/quote]But what percentage of the time is the wrong decision made based on those subjective criteria?

RE: The Robot DBA

Ben Moorhouse — Wed, 07 Sep 2011 08:17:24 GMT

Surely though if everything is being run by Manna, it will be able to use it's own metrics to decide which group of people fit best? It's not requiring human interaction to make the deal anymore.

RE: The Robot DBA

majorbloodnock — Wed, 07 Sep 2011 08:09:17 GMT

If there is a task that can be automated, then it probably will be sooner or later. However, the value of a person in a role is not their ability to do predefined tasks; it's their ability to deal with the subjective.I'll admit I haven't read the story, but it's understandable that a piece of software could organise staff levels and task priorities because optimums can be measured and quantified. What is less believable is the automation of contract negotiation; the question of "do I want to do business with this company?" can hinge on all sorts of intangible and subjective criteria which are unmeasurable even if a person can instinctively weight them up in the mix.A DBA's or a developer's value is often not his or her encyclopaedic knowledge of their tools, but instead their understanding of how the use of that tool will affect business. That's something that can't be replaced by clever software, and perhaps also underlines the importance of techies not ignoring their business acumen.

RE: The Robot DBA

jcrawf02 — Wed, 07 Sep 2011 08:04:52 GMT

Ideally, using automation to replace workers would just require different and/or higher-level skills. The need for people to do *something* will never go away, the skills required will just change.

RE: The Robot DBA

Sailor — Wed, 07 Sep 2011 06:16:35 GMT

People are being asked to do more than one just job or have more than one skill. I admin Linux AND Windows related to web sites besides two (and at one point three) different databases (SQL, DB2 and Oracle). It's a great exposure to everything, but it gets tough to know it all.

RE: The Robot DBA

IceDread — Wed, 07 Sep 2011 03:50:29 GMT

I am right now developing a product that will make a assignments not require humans anymore. These people still posses good knowledge and can focus and be worth more in other parts of the company, but thus I can see first hand this happening at a small scale.I do believe less and less humans will be needed for different tasks. Which IT tasks such as DB management will be left or not and when change takes place is impossible to see but it is for sure that the future will remove some parts of peoples jobs and perhaps introduce new jobs. Once we do have decent "robots" in place the need for a cleaner will be eliminated. However, it also must be cost effective. The world is sure to change, history can tell us this. When and what will change however is harder to predict, I believe it will take a long time before that story comes true if it ever does while I believe my example of cleaners might come to place in a hundred years or so.

RE: The Robot DBA

Ben Moorhouse — Wed, 07 Sep 2011 02:47:07 GMT

Whilst I only read a short excerpt from the Manna story, it's interesting to see that a project I have worked on in the past scheduled which part of the business (and so which task) staff should be working in throughout their shift in 15 minute blocks, based on customer demand and trends for that branch.It showed the manager how many staff he should need, where they should work and what shift patterns needed to be worked.Although facial recognition was never implemented, there was also talk of cameras enabling facial recognition to track both employees and customer behaviours to further improve efficiency....This sounds remarkably like the infancy of Manna... :crazy:

The Robot DBA

Steve Jones - SSC Editor — Wed, 07 Sep 2011 00:32:35 GMT

Comments posted to this topic are about the item [B]The Robot DBA[/B]