Saturday, June 30, 2012

tenure feabhra 2010

Sunday, February 21, 2010

Discussion document ; a cognitive science center for the West of Ireland

1: Introduction / 2: The institutional framework /3: Social Context/4: The Arts

Executive Summary

The native software industry in Ireland, and with it other initiatives like biocomputing, has reached a crossroads. A cost-effective way forward that creates a degree programme in cognitive science and offshoots in applications is outlined.
Introduction


1: Introduction

Although the failure of Medialab Europe was predictable, and predicted, aspects of the rationale for its establishment were and remain correct. The fears that Ireland would end up with little other than call centres, low corporate tax rates, and an uncertain future in an expanded Europe where basic costs are lower elsewhere have proven justified. A generation’s living standards can be protected only once by passing on the bill in the shape of vastly higher property prices to the next generation, and that has been done.

Ireland would be currently at an economic crossroads even if international investment continued at the current rate of 2003-2004. The ratio of native to foreign industry is in many ways the opposite of Israel’s. There have been attempts like SFI, loosely based on the NSF in the US, to bring state-of-the-art research to Ireland, but they have excluded the NSF’s entrepeneurial wing. In parallel with this, entrepeneurship within the universities has been stifled by a management structure that rolls administrative, academic, and technical issues into a single hierarchy. The understandable response of many Irish academics has been to opt for a quiet and civilised, if underachieving existence, in preference to diving into the crucible of dealing with what frankly are often under-qualified administrators in order to implement initiatives.

The 1997 Universities act has resulted in often ill-equipped university management engaging in truly egregious behaviour to staff and students. Simultaneously, the institution of academic tenure, accepted even in private US universities, has been destroyed as management feel free summarily to dismiss permanent university staff, often wrecking world-class labs in the process, and to use the resources of state to submit endless appeals when they lose in court. Were this ruthless behaviour in some way conducive to efficiency and excellence, some kind of case, however ill-motivated, could perhaps be made for it. What has emerged instead is irredeemable mediocrity at best, and a situation where normal academic freedom, a venerable institution in every other Western country, is stifled.. While the 1997 act is thoroughly overhauled, a new model is necessary.

The generally profoundly low ratings achieved by Irish universities in the recent international surveys sit uneasily with the impetus of the 1997 universities act in removing many aspects of what are state universities from the control of state, and vesting frightening concentrations of power in university presidencies. Effectively, what has happened is privatisations without the benefits of exposure to market forces. This is precisely the opposite to Stanford, a private university – and one of the world’s finest – that resembles a coalition of small businesses with little top-down control. Thus, in founding Google, the entrepreneurs Larry Page and Sergei Brin capitalized on the world-class educational and technical expertise at Stanford, but were free to remove themselves from its control, and IP claims. The response of Stanford’s president, John Hennessy, has been personally to buy Google stock.

We propose an entirely new model of academia, entrepeneurship, and research implemented in an intersecting set of structures. While the specifics of the model make reference to particular locations, the founding impulses are much deeper in origin. We propose that an IT in Clare be recognised by the Irish state, and that it be financed as a private-public initiative. The IT should initially run a variety of courses, described below, that are currently not catered for by the state, and should seek university-level accreditation for them in Ireland or elsewhere. The initial courses should service a research and development centre analogously to how the symbolic systems program serves Stanford. They will initially be taught over the web.

In parallel with that, and as recognition of the parlous state of the Irish software industry, it should offer highly subsidised office space for promising start-ups. We believe that this is the type of innovative state involvement that, to take but one example, Peter Evans praises in his analysis of the IT industry in Brazil, Korea, and India (“Embedded autonomy: states and industrial transformation”).

For the next section, we shall outline what cognitive science is, its importance to the knowledge economy, and the courses initially to be run over the web, and then in the College in Clare. While the buildings are being commissioned, a graduate wing can be set up at short notice. Moreover, the courses can be put on the web at short notice; as a proof of concept, we uploaded part of a course in neuroscience taught at Stanford in macromedia flash.


Cognitive Science

Cognitive Science attempts to describe the mind as an informational system. It brings the scientific method to bear on problems of philosophy, psychology and artificial intelligence which have recently become empirically tractable with advances in modern brain imaging technology and the growth of previously unimagined computational resources. It is agnostic as to whether this description captures all aspects of mind; for example, higher-order consciousness may be outside its purview.
It is intimately tied to many previously disparate fields. As such it is in a position to promote interaction between fields and individuals who otherwise would not be in contact. The benefits of such interaction and the quality of academic work which results is in evidence in the proceedings of the international conferences which we have hosted every year from 1992 to 1999, when we bowed to the inevitable( in that we were already experiencing enormous obstruction from the state) represented by the government’s decision in setting up Medialab by suspending the conferences.

Moreover, the two conferences involved, CSNLP and Mind, have successfully been transferred to other locations, like Sheffield and Galway. These conferences, which were self-funding non-profit ventures, brought over a million euro into the state from the attendees.

John Benjamins and prestigious journals have accepted all proceedings we've offered for book publication. The calls for the conferences always originate from Nous Research and the Cognitive Science Society of Ireland(CSSI).

In the context of academia Cognitive Science is positioned to have its findings feed directly into many areas, including applied engineering. Already it has provided the bedrock research which has led to the development of artificial neural networks, expert systems, speech recognition systems, natural language processors, virtual reality interfaces and much more. Two early spin-offs from our own activity include a neural network tool which has been used by developers of cochlear implants (The Brain Construction Kit), and a speech recognition system (Xvoice) which has gained widespread use among quadriplegics and sufferers of repetitive strain injuries.

The next generation of researchers will not simply be designing new and smarter machines, as their forebears did. On a theoretical level, they will ask the question, "What is it to be human?" At the implementation level their work will involve systems which blur the distinction between man and machine. The ethical implications of such work, and of the pursuit of scientific knowledge in general are obvious. What is certain, however, is that a new set of technologies are emerging which emulate aspects of cognitive functioning. The first such are the range of decision support systems, often based on "neural networks", and language tools like speech interfaces and machine translation aids that currently obtain.

The recent emergence of systems which directly capture neural impulses must, however, give pause. Is it possible that we are entering an era of more direct human-machine symbiosis for which this successful attempt directly to interface neural signals with the movement of a cursor is a harbinger? The ethical issues involved require conscious engagement. However, it is also important to note that cognitive science, with its emphasis on mentation, offers an opportunity for a dialogue involving science, the arts, and consciousness studies.

The questions asked are weighty, and it is less than expedient to scrutinise them from a single viewpoint. As such the contribution of artists, psychologists, philosophers and those with backgrounds in the humanities in general will be drawn upon to bolster and inform the scientific process. Practitioners will need a broad skill set consisting of technical virtuosity coupled with an understanding of the subtleties of behavioural psychology, philosophy and a grounding in the social sciences.

The pace of technological advancement makes the delineation of technical skill sets a full time task; however, while the minutiae constantly change and require monitoring, the basic skills remain the same. A firm founding in mathematics and logic provides the individual with the general basis for a proper appreciation of both hardware and software. The details of the electronic engineering underpinning the computational explosion of recent years must be imparted. The ability to provide quality software solutions to immensely complex problems is an absolute essential for success in the field, and we will equip our students with same.

Over and above traditional technical education, we endeavour to engage the individual's sense of self and spirit. If we are to study the human mind, we must each begin with our own, and while it may be the case that emotive instincts are epiphenomenal anomalies, we allow for the reality of the subjective, and encourage its exploration. The relationship between science and society naturally follows from such activity, and will be considered at length.

Cognitive Science has roots in and consequences for philosophy, psychology, linguistics, neuroscience, AI, anthropology, paramedical studies, engineering and biocomputing.

The ACL degree at DCU, established by the founder of Nous Research, provides a precedent for a more general Cognitive Science undergraduate degree, which will have complementary specialisations in cognitive neuroscience and applied intelligent systems, the latter which could be termed “symbolic systems” in keeping with the Stanford nomenclature. Regrettably, the well-documented collapse of the ACL degree points to the precise shortcomings in university structures alluded to above. Remarkably, the founder of the degree was not consulted on any of the appointments made; these have cost the State dearly. (To be precise, around E20 million; 8 permanent staff with 30 years of their contracts to go, infrastructure, and so on.)

The cognitive neuroscience stream, related to the cognitive science program at UC Berkeley, will concern itself with intelligence as manifested in biological systems and the modelling of such systems in computational terms. Parallel distributed processing architectures will be explored at length, as will the details of electrochemical and cytoskeletal processing in the central nervous system. This burgeoning area has many applications from control systems technology to prosthetic implant development.

The applied intelligent systems stream, a counterpart to the symbolic systems program at Stanford, will concern itself with human-computer interaction, robotics and artificial intelligence in general. On a more theoretical front, it is here that we will explore the implications of embodiment for the problem of intelligent machine behaviours. Development of effective artificial perceptual systems in such an environment will present many challenges, but has and will yield many insights into human perceptual structures in addition to the obvious applications of successful models in industry.

For the remainder of this short paper, we shall explore the institutional framework proposed, some further philosophical background, and the consequences for the arts.



2. The institutional framework


There is a danger that, by focusing on programs which develop only specific skills, Ireland will soon allow herself be outbid by countries like India for foreign computing investment. It is imperative that we also explore areas in which we can become true world leaders in the next few decades. We need to focus on just well-chosen areas of excellence

We made a good start in multimedia; at one point, Ireland commanded 40% of the total EU funding from the Telematics program. Companies like Motorola, Intel, Ericsson and Nokia are actively researching how to include the next stage, intelligent multimedia, in their products. Briefly, this requires that such products should have models of the needs and plans of their users. Consequently, intelligent multimedia research is a creative activity involving not just standard software engineering, but also expertise in artificial intelligence, cognitive science, user interfaces and so on.

In a trial run, we set up an independent college. (Negotiations about accreditation at CEO level with the HEA were stalled in 2003 as other problems emerged for the HEA, and in summer 2006 we cut a deal with the HEA to allow collaboration in Ennis with other Irish educational institutions. Accreditation and incorporation can also be done in the US)

One of the courses taught at Erigena College has become a standard in the international consciousness research community; two others have been accredited at Stanford and one other at UC Berkeley. There is no proposal to relocate staff currently in other faculties within the centre. Moreover, once its world reputation in intellimedia is secured, it will have built in to its staffing structure sufficient flexibility to move on to other areas, as necessary. It is proposed to start immediately with a core staff of ten, and build naturally with respect to numbers on the courses.

3. Social context

It is indeed possible that many perennial philosophical problems will find themselves restated and perhaps solved due to current scientific research. An obvious one is the extent to which human knowledge is innate; recent scientific research is ever finding further evidence for innate knowledge. The Centre for cognitive science will be alert to such findings. Moreover, a glance at the conferences we've run will show more than a passing concern with the ethical.

4. Cognitive Science and the Arts


One cannot explore human cognition without being drawn to its finest manifestations, in the past education placed great emphasis on endowing a capacity for both the scientific and the artistic, and we are beginning to rediscover an appreciation for the wisdom of this in more recent times. The dual acts of perception and production of music provide an example of an increasingly active subject of research. Our 1999 conference "Language, Vision and Music" (published as a book by Benjamins) attracted a wide range of perspectives ranging from performing artists concerned with the emotive aspects of their activity to mathematicians intrigued by the implicit structure they have discovered in the language of music.

Cognitive scientists are somewhere in the middle, mediating the two approaches much as the perceptual system itself mediates between structured ontological reality and our subjective experience of it. This does not imply that anything less than the most rigorous scientific procedures are brought to bear upon the study of the arts, but we do seek to expand the territory of our inquiry, and are encouraged by the fruits borne of open-mindedness thus far.

That artistic performance endows the individual with heightened awareness of physical surroundings, rhythm and timing, and most interestingly inter-individual emotive power, is undeniable. As such, we have already established contact with people actively engaged in the performing arts at a high level of accomplishment, that we might better glean an understanding of the processes involved in the fine-tuning of the motor and perceptual processes to their highest degree. While we feel that there is great scientific utility to be obtained by a concern for artistic activity, we also feel that it is a worthy end in itself.

In particular, the now relatively moribund Irish music industry will benefit greatly from its many able participants getting a technological fillip. Cognitive science is primarily the study of human cognition, and therefore the activity of humans themselves.

In short, then, County Clare deserves and needs an autonomous third-level institution. The demographics are against anything resembling a conventional university or polytechnic. We believe that what Clare can realistically expect to get initially is an institution with strong ties to R+D that has postgraduate programs accredited by outside agencies. We have made a start on all of this, and believe that, with government support – or, at least, non-interference – we can rejuvenate parts of the West of Ireland. In 2005, the relevant elected officials in Clare voted for this proposal, and we look forward to its seeing the light of day.

Seán Ó Nualláin Ph.D 21u Feabhra 2010

PS I cannot help noting Prondzynski's personal financial links with skilsoft, already exposed in "the Phoenix" (with an amplification apparently forthcoming, and reprinted below), and wondering if this explains many of his (on the surface) mad decisions; he simply wanted to put money in his own pocket, at whatever cost to the Irish taxpayer(24/2/10)

27/2/10

While outlining his renumeration for Skillsoft,Forbes refers to other unnamed business interests;

http://people.forbes.com/profile/ferdinand-von-prondzynski/72367

It would be interesting to see how much DCU research in the form of students projects were directed to skillsoft and other tasks, marginalising the cognitive science proposal he was shown in Jan 2001

The Phoenix article (Feb 2010)

"As Goldhawk has previously noted (see The Phoenix 18/12/09), Ferdie has a huge appetite for communicating even the most minor details online, whether via Facebook, his blog or, increasingly, on Twitter. But, while his followers were recently informed, for example, that Neil Jordan had stopped to have a chat with fans and sign autographs at Montrose while Colin Farrell kept moving, there was no mention of his good news concerning the Skillsoft, a Nasdaq-quoted e- learning company (founded by Bill McCabe as SmartForce) where he sits on the board. Apart from his fees here, Von Prondzynski was also allotted substantial share options and the good news is that the company is now being taken over at a premium. The buyers are three private equity groups – Berkshire Partners, Advent International and Bain Capital – who have made a $1.1bn all-cash offer which, at $10.80 per ordinary share, represents a 26% premium above the average Skillsoft closing price over the last year and indeed, a 50% premium to the average closing price over the previous five years. Von Prondzynski was on the committee of independent directors who recommended the deal earlier this month. Happily for the DCU president, at the end of January last he held 10,000 shares, now worth $108,000, plus a further 105,000 share options which would be worth big bucks depending on how many he has cashed in. Von Prondzynski’s other interests include a large share in a castle outside Mullingar Co Westmeath – Knockdrin Castle, where the balance is held by his family’s Isle of Man vehicle, Francmine Ltd. "
0 comments