Professional Issues in Computing Computers and Society The

Professional Issues in Computing Computers and Society The Internet

Technology and Terminology n n www. - A virtual network running on the Internet, using hypertext technology. Browsers - Programs which enable users to search [*] through hypermedia, using links. Home page(s). Portions of hypermedia open to general access via the Internet. Note: None of this was new technology - It represented a new combination of existing technologies

History 1 n n n First interlinking of computers (4) -1969 ARPA/DARPA - 1970's ARPANET - 1971 Recent Internet growth 100% per 10 months Future Internet growth predicted to be even greater (health warnings).

History 2 n n n Modem Archie Anonymous FTP Gopher Veronica/Jughead

History 3 n Gutenberg (1453)

History 3 n n n Gutenberg (1453) Music(1465) Caxton (1474)

History 3 n n Gutenberg (1453) Music(1465) Caxton (1474) De Impressione Liborum (1515)

History 3 n n n Gutenberg (1453) Music(1465) Caxton (1474) De Impressione Liborum (1515) Luther’s 95 Theses (1517) Luther’s Bible (100 k copies) (1522)

History 3 n n n n Gutenberg (1453) Music(1465) Caxton (1474) De Impressione Liborum (1515) Luther’s 95 Theses (1517) Luther’s Bible (100 k copies) (1522) 1530 s -1540 s reformations in Scandinavia, France, and Britain. Leading to vast upheaval, start of modern capitalism etc.

History 3 n n n Beware ‘single-cause’ explanations of history. Other things were happening at this time: Glasgow University (1451) European ‘discovery’ of America (1492) Spanish Inquisition (1492) Coffee introduced into Europe (1517)

Technological Issues n n n The difficulty of sensible prediction. No obvious obstacles to improvements and optimization. 'M 25 Syndrome'. Improvements to browsers: 'Intelligent Agents' 'Knowledge-based filters' 'Broadcatch replaces broadcast' Greater bandwidth therefore : Full-motion video 3 D/VR? 'Telepresence'?

Social Issues n 'Information overload' n - technological solutions? - further social implications? The 'information rich' and the 'information poor'. New social groupings? : - nation states - multinational organizations - by interests - by attitude n n n n n Is this a threat to existing power structures? This depends on a wide range of constraints - some of them are being decided now.

Social Issues n n n Allowing the free flow of information is a threat to those who hold power by controlling that flow. Many organizations would like to control the flow of information between individuals and across national boundaries for a wide range of motives. We probably approve of motives such as reducing international crime, terrorism, and preventing the circulation of paedophillic material. We might not approve of the suppression of criticism of governments, particularly if those governments merit criticism. Consider also the consequences of allowing widespread control by: a single media owner; a single software company; a single religious organization; or a relatively small consortium of commercial interests.

Social Issues On the other hand. . . n n n n No control leaves us many problems. including: Use of www. for totally immoral purposes. Use of www. for deceiving consumers in order to gain commercial advantage. Use of www. to spread false information for various purposes. Use of www. to avoid, copyright, intellectual property legislation, and existing censorship. (t. i. n. a. e. l. )

Control of the Internet Technical Issues n n n Is it technically possible? Would damage limitation policing be enough? C. D. A. British methods of control. Unregulated countries. International agreements. Who should do it? Governments? International organizations? Cybercops? Cybervigilantes?

Your role n n n Many of these questions are being decided now. You will need to take a view. “The wild cyber-frontier”

Military, Safety-Critical, and Expert Systems applications

Military Applications of Computing n n n In general the military want computing technology for the same reasons as industry: Communications - bigger, faster, more reliable. Human Replacement - battlefield automation. Decision support - command control (C 3. I. ) We can make some clear distinctions between the implications of these three areas.

Military applications of computing The effects of battlefield automation (eg. smart bombs, unmanned tanks, RPVs) will generally be beneficial. n Improved communications should help reduce mistakes. n C 3. I. on the other hand. . . n Consider the implications of L. O. W. , S. D. I etc. Should we/ are we ready to give up ultimate human control of such things?

Safety-Critical Applications n More computing applications now exist where serious consequences would follow from any mishap or error

Safety-Critical Applications n n n n n More computing applications now exist where serious consequences would follow from any mishap or error - for example: Nuclear power generation. Chemical process control. Flood control and drainage systems. Automated transport systems. Fly/drive-by-wire systems. Medical monitors. Medical advice-giving systems. Automated weapons. Military decision support (C 3. I. )

Safety-Critical Applications n More computing applications now exist where serious consequences would follow from any mishap or error. Some questions:

Safety-Critical Applications n More computing applications now exist where serious consequences would follow from any mishap or error. Some questions: 1. Can we build safe enough systems?

Safety-Critical Applications n More computing applications now exist where serious consequences would follow from any mishap or error. Some questions: 1. Can we build safe enough systems? 2. Who is responsible when things go wrong?

Safety-Critical Applications n More computing applications now exist where serious consequences would follow from any mishap or error. Some questions: 1. Can we build safe enough systems? 2. Who is responsible when things go wrong? 3. How safe is safe?

Safety-critical applications n n n n How safe is safe? Performance to specification. Exhaustive testing. Limit and typical testing. Mathematical or logical validation. Modular design. Modular programming effort. None of these are entirely satisfactory, but so far the software industry has outperformed manual methods in safety-critical applications!

Expert Systems (Hayes-Roth 1984) n They can solve very difficult problems. n They reason heuristically and can interact with humans appropriately (eg. in natural language). n They manipulate and reason about symbolic descriptions. n They can work with data containing errors, using uncertain judgmental rules. n They can contemplate multiple competing hypotheses simultaneously. n They can explain why they are asking a question n They can justify their conclusions.

Expert Systems n n n But : What exactly is their failure mode? Who will take responsibility?

Embedded Technology n Much of what has been talked about in this lecture will be embedded in other types of system in the future. n Expert or Knowledge-based techniques are now general. A. I. techniques are becoming mainstream. What is in the laboratories today will be in the shops tomorrow. This may add significantly to the range of social and moral issues that you need to consider. n n n