POV-Ray : Newsgroups : povray.off-topic : Preparedness Server Time
29 Jul 2024 10:30:41 EDT (-0400)
  Preparedness (Message 103 to 112 of 142)  
<<< Previous 10 Messages Goto Latest 10 Messages Next 10 Messages >>>
From: clipka
Subject: Re: Analysis
Date: 8 Sep 2012 08:21:25
Message: <504b3845$1@news.povray.org>
Am 08.09.2012 05:37, schrieb waggy:
> nemesis <nam### [at] gmailcom> wrote:
>> Em 07/09/2012 08:24, scott escreveu:
>>>>> Manufacturing uses robotics in a huge way.
>>>>
>>>> OK. But those robots already exist. Why would you ever need to design
>>>> more?
>>>
>>> Competition. If you don't design new robots that are faster, more
>>> accurate, more versatile and cheaper than your previous generation, your
>>> customers will buy their robots from someone who has designed new ones.
>>
>> at least until robots do the design of faster and better robots.  Then
>> no one will do any more shopping.
>
> So the Robot Apocalypse ends suddenly when they all decide it would be more fun
> to just go shopping?

http://xkcd.com/1046/


Post a reply to this message

From: Stephen
Subject: Re: Analysis
Date: 8 Sep 2012 09:00:42
Message: <504b417a$1@news.povray.org>
On 08/09/2012 1:21 PM, clipka wrote:
> Am 08.09.2012 05:37, schrieb waggy:

>>
>> So the Robot Apocalypse ends suddenly when they all decide it would be
>> more fun
>> to just go shopping?
>
> http://xkcd.com/1046/
>

Andrew? ??? ;-)

-- 
Regards
     Stephen


Post a reply to this message

From: Darren New
Subject: Re: Analysis
Date: 8 Sep 2012 11:41:40
Message: <504b6734@news.povray.org>
On 9/6/2012 2:34, Orchid Win7 v1 wrote:
> rather than "web-based business development".

This is secret code for "10+ hour days trying to meet the unrealistic 
expectations and deadlines of our clients, rather than someone who you can 
actually talk to when they tell you to do something infeasible."

-- 
Darren New, San Diego CA, USA (PST)
   "They're the 1-800-#-GORILA of the telecom business."


Post a reply to this message

From: Darren New
Subject: Re: Analysis
Date: 8 Sep 2012 11:44:48
Message: <504b67f0$1@news.povray.org>
On 9/6/2012 9:52, Orchid Win7 v1 wrote:
> So apparently there were a shedload of people far better than me
> to choose from.

You can't rank something like that on a linear scale. There were a number of 
people they thought were better for their job than you were, yes.

-- 
Darren New, San Diego CA, USA (PST)
   "They're the 1-800-#-GORILA of the telecom business."


Post a reply to this message

From: Darren New
Subject: Re: Analysis
Date: 8 Sep 2012 11:46:30
Message: <504b6856$1@news.povray.org>
On 9/6/2012 9:46, Orchid Win7 v1 wrote:
> The thing is, I can't think of any career that makes use of an ability to
> memorise and apply pointless algorithms...

The legal system. It's like hacking English.

-- 
Darren New, San Diego CA, USA (PST)
   "They're the 1-800-#-GORILA of the telecom business."


Post a reply to this message

From: Darren New
Subject: Re: Analysis
Date: 8 Sep 2012 12:04:57
Message: <504b6ca9$1@news.povray.org>
On 9/6/2012 12:05, Orchid Win7 v1 wrote:
>>> The thing is, I can't think of any career that makes use of an ability
>>> to memorise and apply pointless algorithms...
>>
>> Off the top of my head:
>
>> research
>
> Seriously? That's a career? Talk about "vague"...

The point is that whatever you research in any field, you're going to need 
that ability.

>> finance
> Wouldn't I need extensive finance qualifications to do that?

No. The finance guys have extensive knowledge. They rely on technical guys 
to implement their algorithms. I know several mathematicians who don't know 
how to predict stocks, but know how to program computers with algorithms to 
predict stocks. And they use baroque custom languages like R, K, etc.

>> crypto/security
> Seriously, almost *nobody* actually does that.

Really? Wow. I guess all the people working in that field would disagree. 
You realize this includes things like analyzing how you can tell someone is 
breaking into your system and such, right?

>> logistics
>
> Really? I'm sure it's difficult keeping track of stuff, but I don't see much
> technical detail involved. It's just about keeping track of a lot of stuff
> all at once. It's hard, but it isn't very technical.

It's about getting it there efficiently. Logistics: Customer orders four 
dump trucks of cement. You have to figure out when to send the trucks to get 
it there soon enough after they're ready that they don't waste time, but not 
so soon that it congeals inside the truck. You also have to know how they're 
going to use it, so you can space out the trucks so the second is ready 
before the first has set enough. You have to have real-time traffic 
estimates, knowing how to route around them, including rerouting trucks 
already moving to get them where they're supposed to go at the right times. 
You have to account for days when a truck breaks down, a driver is out sick, 
or a construction company cancels or defers an order.  And that's just local 
delivery of fucking *cement*. What do you think a company like Amazon or UPS 
needs?

>> data mining
> Does anybody actually do that? I thought it was just a fashionable middle
> management buzzword.

That's ... all google does, basically. Every company that actually has a 
successful business needs constant "Actionable Business Information." How 
much to order. Who to hire. Where to advertise.

Heck, I write computer programs at Google, and my salary is based on how 
much money those computer programs make, just like salesmen get commission. 
How do you measure it? Well, if the user watches a lesson on my program, and 
then goes into one of the other programs and does what the lesson told him 
to do, and from that Google makes more money, then I attribute part of that 
to my program. *figuring out* those relationships is difficult. Going the 
extra step of looking at what he's already doing and suggesting other things 
he could tell his clients to ask for is yet another level harder.

Not that it's especially complex math, but the complexity is pretty 
astounding, especially when you're talking about the size of databases 
Google deals with.

>> robotics
> Is there any commercial application for that?

http://www.youtube.com/watch?v=kvf29R7nXlM

>  From what I've seen, installing comms equipment doesn't require a great
> deal of technical skill - you just need to be good at carrying stuff around
> and running cables.

Heh heh heh. So, you already know how to calibrate a cell phone tower 
installation? How to splice optical fiber? How to run 10Kilovolt power 
circuits?

You've only installed stuff that was built for retail installation, clearly.

Granted, probably not a whole lot of programming involved.

>> marketing
> There's technical expertise in that?

If you're marketing technical products, you'll sell 10x as much if the 
person writing the ad actually knows what the product does, yes.

Do you think you could talk to the manufacturer and write a reasonable ad 
for a desktop computer system or a tape backup carousel? Do you think you 
could talk to the manufacturer and write a reasonable ad for an 
anti-retrovirus drug?

>> disaster response
> What kind of disaster response requires technical skill?

The kind where the disaster has struck technical infrastructure. Didn't you 
write the disaster recovery documents for your previous job? Could you have 
done that without knowing anything about technology?

-- 
Darren New, San Diego CA, USA (PST)
   "They're the 1-800-#-GORILA of the telecom business."


Post a reply to this message

From: Darren New
Subject: Re: Analysis
Date: 8 Sep 2012 12:13:49
Message: <504b6ebd@news.povray.org>
On 9/7/2012 3:24, Orchid Win7 v1 wrote:
> Ironically, I'm really quite bad with numbers. I'm good with /equations/...

That's what finance is about.

> You're telling me there are more than three people on Earth who actually
> design ciphers?

There's a lot more to security than designing the cyphers. And yes, Bellcore 
had, for example, a 250 person department doing just that sort of stuff. 
Ever hear of the NSA? Is it MI6 where you are, or some other part?

> I would have thought that querying the data to get the numbers you want is
> the /easy/ part.

It's really not, because the numbers aren't in one place. The number you 
want is "what percentage of what I spent on each project (ad, program, etc) 
actually turned into profit money?"

> The hard part, surely, is figuring out what questions to
> ask in the first place. And that is out of my league.

Nope. The easy part is asking the question.  "If I buy the new lab analysis 
machine for $100K, how long before I make that money back, and how much more 
will I make over its lifetime?"

Or "how frequently should we make backups? If we make them too frequently, 
then the amount of effort we spend making backups we never restore outweighs 
the amount of effort it would take to reproduce the results from the time of 
the last back up to the time of the crash where we needed the restoration."

> OK. But those robots already exist. Why would you ever need to design more?

The robotics for making space stations don't really exist. Nor do the 
robotics for manufacturing flying automobiles.

>>>> marketing
>>>
>>> There's technical expertise in that?
>>
>> Yes. Writing marketing copy that actually convinces technical people
>> requires technical expertise.
>
> You must be looking at very different "marketing copy" than the stuff I've
> seen.

That's because you're not looking at marketing copy for products marketed to 
technical people. Look at some ads for computers on Amazon.

> In particular, such material is utterly devoid of even the slightest hint of
> technical detail. Lots of hand-waving about "total cost of ownership" and
> "return on investment" and so forth, but no technical specifications, and no
> prices.

Because that's the numbers the business people want to know. See above.

Now, go look at an ad for a scuba computer. Or a camera.

> Or rather, it requires technical skill to design the data center correctly
> in the first place. By the time a disaster actually occurs, it should be
> easy enough that a trained monkey could do the actual recovery part...

It depends how big your data center is. Trust me, when something "goes 
wrong" at a google data center, and they start seeing 15-second drops of 4% 
packet loss every hour or so over various inter-center data connections, 
figuring out wtf is going on is quite technical.

-- 
Darren New, San Diego CA, USA (PST)
   "They're the 1-800-#-GORILA of the telecom business."


Post a reply to this message

From: Darren New
Subject: Re: Analysis
Date: 8 Sep 2012 12:15:38
Message: <504b6f2a$1@news.povray.org>
On 9/7/2012 8:48, Jim Henderson wrote:
> That's not actually how it works.  I know from having worked on DR plans
> for large and small companies.

Indeed, if it worked that way, you would have just planned to avoid the 
disaster in the first place.

-- 
Darren New, San Diego CA, USA (PST)
   "They're the 1-800-#-GORILA of the telecom business."


Post a reply to this message

From: Orchid Win7 v1
Subject: Re: Analysis
Date: 8 Sep 2012 12:25:00
Message: <504b715c$1@news.povray.org>
On 08/09/2012 04:46 PM, Darren New wrote:
> On 9/6/2012 9:46, Orchid Win7 v1 wrote:
>> The thing is, I can't think of any career that makes use of an ability to
>> memorise and apply pointless algorithms...
>
> The legal system. It's like hacking English.

Not really. Computer systems follow rules. The legal system does not. 
(And neither does English, BTW.)


Post a reply to this message

From: Orchid Win7 v1
Subject: Re: Analysis
Date: 8 Sep 2012 12:30:35
Message: <504b72ab$1@news.povray.org>
>>> finance
>> Wouldn't I need extensive finance qualifications to do that?
>
> No. The finance guys have extensive knowledge. They rely on technical
> guys to implement their algorithms. I know several mathematicians who
> don't know how to predict stocks, but know how to program computers with
> algorithms to predict stocks. And they use baroque custom languages like
> R, K, etc.

I saw an interesting comment on LinkedIn about this - saying that nobody 
uses Haskell for this kind of thing because it's "non-deterministic", in 
that the calculation might take 10 seconds or 10 years, and there's no 
way to know which. [This is obviously false, but I can see where 
somebody might get that misconception...]

> It's about getting it there efficiently. Logistics: Customer orders four
> dump trucks of cement. You have to figure out when to send the trucks to
> get it there soon enough after they're ready that they don't waste time,
> but not so soon that it congeals inside the truck. You also have to know
> how they're going to use it, so you can space out the trucks so the
> second is ready before the first has set enough. You have to have
> real-time traffic estimates, knowing how to route around them, including
> rerouting trucks already moving to get them where they're supposed to go
> at the right times. You have to account for days when a truck breaks
> down, a driver is out sick, or a construction company cancels or defers
> an order. And that's just local delivery of fucking *cement*. What do
> you think a company like Amazon or UPS needs?

Hmm, interesting. I assumed they just get it wrong most of the time.

>>> data mining
>> Does anybody actually do that? I thought it was just a fashionable middle
>> management buzzword.
>
> That's ... all google does, basically.

Yeah, I'm still trying to wrap my mind around how Google actually 
obtains any income...

> Not that it's especially complex math, but the complexity is pretty
> astounding, especially when you're talking about the size of databases
> Google deals with.

OK. But does any other company on the planet do that?

>> From what I've seen, installing comms equipment doesn't require a great
>> deal of technical skill - you just need to be good at carrying stuff
>> around and running cables.
>
> Heh heh heh. So, you already know how to calibrate a cell phone tower
> installation? How to splice optical fiber? How to run 10Kilovolt power
> circuits?
>
> You've only installed stuff that was built for retail installation,
> clearly.
>
> Granted, probably not a whole lot of programming involved.

I've seen a guy splice fiber. He cut the two ends, put them in a 
machine, and pressed a button. It used a digital microscope to align the 
ends in three dimensions, and then weld them together. Freaky as hell. 
Apparently it costs about £8,000. (Which is presumably peanuts compared 
to what the fiber it splices costs...)


Post a reply to this message

<<< Previous 10 Messages Goto Latest 10 Messages Next 10 Messages >>>

Copyright 2003-2023 Persistence of Vision Raytracer Pty. Ltd.