SERVING THE QUANTITATIVE FINANCE COMMUNITY

 
User avatar
Cuchulainn
Topic Author
Posts: 59982
Joined: July 16th, 2004, 7:38 am
Location: Amsterdam
Contact:

Re: Three years from now: can you predict the Software Landscape in anno 2018?

June 16th, 2017, 2:46 pm

Banks in Hong Kong are setting the bar high when recruiting AI candidates and are struggling to attract enough people. They want a Masters or PhD in statistics, computer science or maths; experience in machine learning; knowledge of data science; and skills in C++, Python, R, KDB or MATLAB, says Warwick Pearmund, an associate director at headhunters Harvey Nash in Hong Kong. “Banks prefer people with financial industry experience, but in reality there’s only a small pool to draw from, even globally,” he adds. “And AI candidates are just as likely to opt for a tech giant or boutique trading firm, which often offer more nimble, more relaxed environments.”
 
User avatar
Cuchulainn
Topic Author
Posts: 59982
Joined: July 16th, 2004, 7:38 am
Location: Amsterdam
Contact:

Re: Three years from now: can you predict the Software Landscape in anno 2018?

June 18th, 2017, 10:33 pm

About 15 years ago Agent Technology was a hot topic. Whatever happened?

https://en.wikipedia.org/wiki/Software_agent


Agents are the next major computing paradigm and will be pervasive in every market by the year 2000. (Janca, 1995)
Image
What the graph units, quantities etc.? What is growth?
 
User avatar
Cuchulainn
Topic Author
Posts: 59982
Joined: July 16th, 2004, 7:38 am
Location: Amsterdam
Contact:

Re: Three years from now: can you predict the Software Landscape in anno 2018?

June 18th, 2017, 10:34 pm

About 15 years ago Agent Technology was a hot topic. Whatever happened?

https://en.wikipedia.org/wiki/Software_agent


Agents are the next major computing paradigm and will be pervasive in every market by the year 2000. (Janca, 1995)
Speak of the devil! Agents are back.
https://venturebeat.com/2017/06/17/10-p ... nt-design/

They were also saying this 20 years ago.
It’s a great time to start laying out practices and principles for how we want to design and build intelligent agents. As part of that thought process, below are 10 principles that can help govern the future design of intelligent agents.

http://www.datasim.nl/Education/Special/Agent.asp
 
User avatar
Cuchulainn
Topic Author
Posts: 59982
Joined: July 16th, 2004, 7:38 am
Location: Amsterdam
Contact:

Re: FIVE years from now: can you predict the Software Landscape in anno 2023?

March 1st, 2018, 12:20 pm

Crystal Ball Time, what has the Future in store?

At the moment of writing it is safe to say that the “the dominant paradigm of software development provides little insight into how to build software effectively and it is based on the individual programmer model similar to other information-intensive activities such as novel writing, theorem proving and music composition” (Cox (1996)).  This approach leads to software products that tend not to work well with other software components, possibly written by other developers in the same organisation or by third-party commercial software vendors. Two major shortcomings with current practice is that first the vocabulary in development teams seems to be restricted to words at programming level or in some cases to software design patterns (as discussed in GOF 1995) and second a lack of interoperability between software components.  More generally, “current software practice does not match the usual expectations of an engineering discipline” (Shaw and Garlan (1996)) mainly due to the lack of architectural models and to the lack of interface standardization between software component (Leavens and Sitarman (2000)).  This desire for interface standardization is hampered to some  extent by traditional object-oriented technology. Using delegates in .NET instead of subtype polymorphism results in more flexible software.
Last edited by Cuchulainn on March 1st, 2018, 12:52 pm, edited 1 time in total.
 
User avatar
Cuchulainn
Topic Author
Posts: 59982
Joined: July 16th, 2004, 7:38 am
Location: Amsterdam
Contact:

Re: FIVE years from now: can you predict the Software Landscape in anno 2023?

March 1st, 2018, 12:34 pm

Some of the principles underlying our design approach can be summarised by the steps that György Pólya describes when solving a mathematical problem (Pólya (1990)):
1.      First, you have to understand the problem.
2.      After understanding what you are trying to solve, make a plan.
3.      Carry out the plan.
4.      Look back at your work. How could it be better?
It is interesting to ask how many software projects started life at step 3!!!
 
User avatar
Traden4Alpha
Posts: 23951
Joined: September 20th, 2002, 8:30 pm

Re: FIVE years from now: can you predict the Software Landscape in anno 2023?

March 1st, 2018, 2:14 pm

Some of the principles underlying our design approach can be summarised by the steps that György Pólya describes when solving a mathematical problem (Pólya (1990)):
1.      First, you have to understand the problem.
2.      After understanding what you are trying to solve, make a plan.
3.      Carry out the plan.
4.      Look back at your work. How could it be better?
It is interesting to ask how many software projects started life at step 3!!!
Yes, the spiral model makes sense for simple systems but it feels very reductionist, presumptive of the pre-existence of certain kinds of knowledge, and very slow.

For more novel systems, step 1 may be impossible without first doing some step 3 & 4. If the programmer asks "what do you want" which prompts the customer to ask "what can you do," then it may be necessary to do some quick programming (steps 3 & 4) to even begin to understand the problem/solution space (for both programmer & customer).

And for more fast-paced environments, the linear sequence of 1 through 4 may be too slow. I'm reminded of a problem the US Department of Defense started facing in the late 1980s when PCs came on the scene. The DoD was so used to very careful step-by-step planning for procurement that by the time they approved a contract to buy new technology, changes in technology had invalidated the results of step 1. Japan's fifth generation computer project also exemplifies the failings of over-planning.

The 4-step method does not work very well if the customers are changing at the same time the hardware is changing and the operating systems are changing and the applications are changing. For some categories of problems & solutions, the required development process may be like the amoeba method of gradient ascent with very large numbers of short-range trials. Sometimes the best way to figure out "how" is to simply "do".
 
User avatar
Cuchulainn
Topic Author
Posts: 59982
Joined: July 16th, 2004, 7:38 am
Location: Amsterdam
Contact:

Re: FIVE years from now: can you predict the Software Landscape in anno 2023?

March 1st, 2018, 2:27 pm

The 4-step method does not work very well if the customers are changing at the same time the hardware is changing and the operating systems are changing and the applications are changing. 

Tell us what does work well! Historical anecdotes aside. 

.. similar to other information-intensive activities such as novel writing, theorem proving and music composition
 
User avatar
Traden4Alpha
Posts: 23951
Joined: September 20th, 2002, 8:30 pm

Re: FIVE years from now: can you predict the Software Landscape in anno 2023?

March 1st, 2018, 2:36 pm

The 4-step method does not work very well if the customers are changing at the same time the hardware is changing and the operating systems are changing and the applications are changing. 

Tell us what does work well! Historical anecdotes aside.
Actually, nothing works well!

Programming is a lot like home remodeling: you don't know what you're going to discover and you don't really know what you want until to get 50% into the project, have used 80% of the budget, and find you are only 25% done.

The overall problem that plagues software engineering is one of irreducible complexity: the knowledge required for a well-formed plan does not exist and the resources required to gain said knowledge may be quite large or unbounded.
 
User avatar
Cuchulainn
Topic Author
Posts: 59982
Joined: July 16th, 2004, 7:38 am
Location: Amsterdam
Contact:

Re: FIVE years from now: can you predict the Software Landscape in anno 2023?

March 1st, 2018, 2:53 pm

Actually, nothing works well!

Can I quote you on that?

Maybe your AI world is in constant motion (why?) but there are many well-defined domains that are less sensitive to the chaos you seem to be describing.

And these days there is OpenCl and so on.
 
User avatar
Traden4Alpha
Posts: 23951
Joined: September 20th, 2002, 8:30 pm

Re: FIVE years from now: can you predict the Software Landscape in anno 2023?

March 1st, 2018, 3:53 pm

Actually, nothing works well!

Can I quote you on that?

Maybe you AI world is in constant motion (why?) but there are many well-defined domains that are less sensitive to the chaos you seem to be describing.

And these days there is OpenCl and so on.
It's not just AI, it's almost everything except very siloed, stand-alone applications.

I've got a stack of old computers in the basement. Every one of them is still 100% usable but none of them are useful because so much has changed.

I don't see how OpenCL addresses the challenge of step #1 when designing a new piece of software. Maybe it helps during step 3

-----

The chaos issue is only one half of the problem. Even if every chip, OS, and language were frozen in perpetuity, the four step method still would not work well.

The problem remains that often times you can't do step 1 until you've done some step 3 -- the customer needs to see the software before they can really state what their needs are. That's especially true for commercial software in which the features are cost sensitive.

Does software engineering have accurate tools for predicting the cost and time to develop a piece of software? For hardware, there's a lot of decent cost estimation tools. But if I say I want a customer service application, how much effort does it take to predict whether the project will take exactly 4 FTE weeks versus 5 FTE weeks (which is a 20% overrun)?
 
User avatar
Cuchulainn
Topic Author
Posts: 59982
Joined: July 16th, 2004, 7:38 am
Location: Amsterdam
Contact:

Re: FIVE years from now: can you predict the Software Landscape in anno 2023?

March 1st, 2018, 5:04 pm

I've got a stack of old computers in the basement.

iphone {1,2,...6,7,8} ?
 
User avatar
Traden4Alpha
Posts: 23951
Joined: September 20th, 2002, 8:30 pm

Re: FIVE years from now: can you predict the Software Landscape in anno 2023?

March 1st, 2018, 5:18 pm

I've got a stack of old computers in the basement.

iphone {1,2,...6,7,8} ?
Mac 128k, Mac II, Mac IIsi, Powerbook 190, PowerMac 7500, PowerMac 8500, PowerMac G3, G4 PowerMac, MacPro G5, ... etc. etc. And there's a Sinclair ZX80, various Psions, various Palm PDAs, and other little portables from back in the day.
 
User avatar
Cuchulainn
Topic Author
Posts: 59982
Joined: July 16th, 2004, 7:38 am
Location: Amsterdam
Contact:

Re: FIVE years from now: can you predict the Software Landscape in anno 2023?

March 1st, 2018, 5:23 pm

No Apple II? Not even a wee iphone 1 lurking in a drawer?
When did 'planned obsolescence' raised its head in the industry?
Last edited by Cuchulainn on March 1st, 2018, 5:27 pm, edited 2 times in total.
 
User avatar
Cuchulainn
Topic Author
Posts: 59982
Joined: July 16th, 2004, 7:38 am
Location: Amsterdam
Contact:

Re: FIVE years from now: can you predict the Software Landscape in anno 2023?

March 1st, 2018, 5:24 pm

Does software engineering have accurate tools for predicting the cost and time to develop a piece of software? 

yes, called domain experts.
 
User avatar
Traden4Alpha
Posts: 23951
Joined: September 20th, 2002, 8:30 pm

Re: FIVE years from now: can you predict the Software Landscape in anno 2023?

March 1st, 2018, 5:45 pm

No Apple II? Not even a wee iphone 1 lurking in a drawer?
When did 'planned obsolescence' raised its head in the industry?
During the Apple II era, I was using timeshared DEC10 and CDC6600 machines. My first "computer" was an HP-25 programmable calculator in 1977 (49 steps of assembly language program memory!!!!). Our first iPhone was an iPhone 3 which is lurking in a drawer.

Voluntary 'planned obsolescence' arose in the 1960s with the first appearance of Moore's Law. When a new computer was twice as powerful as a two-year computer, chances are you'd dump the older machine so you could run new, more powerful OSes and apps that required more computing power.

Involuntary 'planned obsolescence' appeared in the 2000s with the 100% adoption of internet connectivity, cybersecurity threats, and forced-march OS & application updating. At this point, any machine older than about 10 years can't be used (it's not interoperable with "modern" systems) and shouldn't be used (it's too vulnerable to malware).
ABOUT WILMOTT

PW by JB

Wilmott.com has been "Serving the Quantitative Finance Community" since 2001. Continued...


Twitter LinkedIn Instagram

JOBS BOARD

JOBS BOARD

Looking for a quant job, risk, algo trading,...? Browse jobs here...


GZIP: On