An academic's view of the cloud?

I was recently interviewed by Computer Weekly during the Beyond the Cloud conference organised by BT and Computer Weekly. The interview is available online.

Barcelona, Interdisciplinary Research and the REF

Blogging has been light - a combination of loads of work to do and then much needed leave to recover!

Right now I'm in Barcelona attending the WCCI 2010 conference where I'm presenting a rather nice algorithm for fuzzy-rough set reduct optimisation. I'll also be working on a journal paper with my collaborators.

What's this to do with information leadership? Well, first of all I like good science and I enjoy working with my co-authors from Aberystwyth University. I also like to get my geek on, once in a while...:-)

The other is more pragmatic. A large chunk of HEFCE income to UK universities (about 20% of a £7476m spend in 2008, see here), is allocated according to a periodic research evaluation process (the one coming up is called the Research Excellence Framework - REF). It also ends up acting as a (contested) measure of relative prestige and appears in some league tables.

Needless to say a >> £1Bn incentive will affect behaviour. Given that teaching is not funded (incentivised) by HEFCE on a 'quality' basis (in fact by student numbers and subject price group), some believe that teaching has suffered - the arguments are well-rehearsed in the HE management literature. The effect of the National Student Survey is at present on reputation (its also worth looking at recent press reports on student contact hours in some research intensive universities).

The problem for setting up a research programme to support information leaders is that something that doesn't neatly fit in a disciplinary silo can suffer. For example, the computer science panel naturally likes its logic, code and algorithms and possibly won't be cognizant of the (intellectual) challenges that information leaders and policy makers face. Other panels will have their own issues, no matter who much they try to be objective (and they do try).

Doubtless high quality mainstream work in business, law etc that the Centre for Information Leadership can stimulate will be recognised. But other core activity of the centre may not necessarily generate results (no matter how high the quality is) that there is a prestigious forum to publish in, or that any single REF panel will fully appreciate. This is a well-known problem for interdisciplinary research, especially those that attempt to look at something new and engage with professions, business and civil society. But such activity is IMHO necessary if universities are to re-energise their relationships with the outside world.

I expect the Centre for Information Leadership to be a game-changer on how universities work with the IT industry, but there's a natural annoyance that we are working in a system that actively incentivises the status quo. The rules of the REF game are fluid and subject to change as the details are argued over. But if you are happy to focus in a core discipline then the fora in which to get quality work recognised by any reasonable version of the REF rules are clear (i.e one knows what they need to do in this game, so a paper in Nature is always good!). Anything else, has to work harder and second guess. A case if one is needed as to how the state can construct unintended dysfunctional systems.

One thing we will do in the Centre for Information Leadership is to find a way of making the research that information leaders need get recognised and funded (its part of my mission - perhaps later blogposts?), but it is prudent to make sure I have good 'core-CS' publications portfolio for the time being. So I'll be publishing good papers on AI algorithms and attending conferences for a while yet.

Computer Science Unplugged - The Show

To start off a series of post on promoting computing in schools, this video shows an entertaining way to introduce computer science to school students. It's given me a few ideas.

For more details of the initiative that gave rise to this see CS Unplugged. Worth a look.



There will be more of these videos posted in the future. Some of them are quite good, others not.

Public Sector CIO Pay: Are we comparing like with like...?

There has been something of a moral panic regarding the pay of some senior civil servants, including some government CIOs getting paid more than the Prime Minister (who appears to have been given a pay cut). I'm not about to make moral or value judgements - this is not the place.

The real issue for me is whether we are comparing like for like. In executive recruitment, you need to consider the whole reward package: pay, perks, use of hotels/residences, memberships, pensions... For example, the PM gets the free use of two 'grace and favour' residences (No. 10 and Chequers) plus associated maintainence, cleaners, servants. I somehow doubt this is matched by the civil service CIOs.

The above applies to making comparisons with private sector peer CIOs too. We also then need to consider the benefits side, i.e. what they deliver. At that point we can start to look at costs and benefits somewhat more objectively, before discussing the value judgements.

All this reminds me of comparisons made routinely in the press about goods being cheaper in the US than in the UK (currently fluctuations aside). These claims don't always stand up to scruitiny as they invariably quote US prices excluding sales tax but UK prices including VAT and/or compare US out of town prices with central London prices. But hey, don't let the facts get in the way of a good story!..:-)

The Recession != Dot-com Crash

Prior to this recession, the (informal) consensus has been that the IT job market is recovering from the dot-com crash whose effects of widespread IT job losses hit a nadir in 2003. Looking at Figure 1 below that looks at the rate of advertised IT positions, it would appear that the recent recession had the same effect.

 

Figure 1: Advertised IT vacancies as of 29 March 2010 (old.jobstats.co.uk)

This is not the case this time round. Unemployment has been felt across the wider economy, and it was the financial services sector that bore the public brunt (although public sector looks like being the vanguard of a second wave after the election). But is the IT profession shedding jobs in large numbers? There have been no high-profile press reports and official unemployment statistics lag behind a dynamic situation, especially when close analysis under the headline figure is needed.

Closer analysis of the available vacancies data paints an unexpectedly positive picture. Figure 2 looks at trends in advertised IT salaries. If widespread IT job losses were in the offing we would expected advertised salaries to fall in response to supply exceeding demand (as was the case in 2003). This has clearly not happened.

 

Figure 2: Advertised IT salaries as of 29 March 2010 (old.jobstats.co.uk)

Closer analysis of the available vacancies data paints an unexpectedly positive picture. Figure 2 looks at trends in advertised IT salaries. If widespread IT job losses were in the offing we would expected advertised salaries to fall in response to supply exceeding demand (as was the case in 2003). This has clearly not happened.

This is not to say that there have not been some job losses due to the recession (e.g. the IT staff at Woolworths), but loses have not been widespread and appear to have be consequential of business failure in other sectors of the economy. The drop in vacancies is most likely due to ‘churn’ being taken out of the job market: employees are not risking job moves, and employers are cautious about hiring. As noted above there are signs this is bottoming out.
 
The recession is therefore qualitatively different for IT professionals, than the dot-com crash. Could IT be seen by the wider world as a relatively safe bet? For sure, the credit bubble has more than wiped out the memory of the much smaller dot-com one! Given the widely-reported problems in recent years of some computing departments being unable to recruit the numbers of students they did at the turn of the millennium this must come as a relief.

Cutting Costs and Public Sector IT

My colleague David Chan, Director of the Centre for Information Leadership has written two challenging articles on cutting costs and public sector IT: Cut deeper - but spend more on ICT and City academic calls for IT cull. We now need to develop and support the transformational information leaders who can deliver.

BP: When an IT failure really isn't one

A number of articles have arisen around the role of IT in the tragic BP oil spill (i.e. BP oil spill ‘slows’ but serious IT failures come to surface and BP oil spill slowing but IT failures revealed).

I have to say I'm rather concerned at these stories in IT publications that I hold in high regard. The reason is that when you read the story and look at the report, it does not back up the claims about IT failure in these articles. They claim that:

"...the US government released a summary of BP’s own early investigation into the problems. The document contains some damning facts about IT at the rig, ..."

Except that it doesn't: the Congressional memorandum makes no mention of IT, or for that matter the words 'information', 'computer' or 'software' (go on, click the above link and see for yourself). The memo does however make plenty of mention of procedural and (non-IT) equipment failures.

There were references to robots and supercomputers in the articles, but in the context of the clear-up (and so are not relevant). The only other part of the two articles that seemed to support their narrative was:

"BP has said the accident “was brought about by the failure of a number of processes, systems and equipment”..."

The problem with leaning on this (unattributed) quote is that it does not specifically apply to IT. In fact the above form of words is so generic that it could apply to anything. The plain English translation would be 'stuff went wrong, we don't know what, but we wish to sound like we do!'.

In short, the content and sources of these articles do not support their title or overall narrative of an IT failure: in fact they point to likely non-IT causes.

Why should this bother me? The reason is that ensuring that an informed public debate on issues as serious as the BP oil leak is important, as is the role of IT in society more generally. Given the blame that will doubt arise as a result of the tragedy, the easy route of blaming IT for the sake of expediency is not the responsible or moral one.