LARGE SCALE COMPUTER SERVICE UTILITIES

CHARTERED ACCOUNTANTS OF ONTARIO. TORONTO, MAY 12, 1969

Where are you likely to see computers next?

A POINT OF VIEW:

Intervisibility map problem – really a complex approach to saying “what you see depends on your point of view”.

Fun to speculate – we will do that later.

But more useful to you to give you a perspective – a ‘do it yourself course’.

Don’t make the mistake of feeling that you cannot contribute some of best ideas come from people who know basics and have imagination.

Consider the computer in its most general sense – extension of man’s mind.

Then think of everything you do in the course of your work in terms of:

• posing problems

• accessing the relevant information

• making logical decisions

you are inviting the assistance of a computer in that work.

THE RECENT PAST

Emphasis in early days was improving batch processing capability.

• main frame speed and reliability

• faster I/O units

• larger on-line storage

This will continue, but in early 60’s more emphasis placed on making things easier for the user.

• have interface

• simulate human approach.

• optional scanners

• visual displays

• voice communication

But as machine got larger and ability to use them improving, more emphasis had to be placed on how to get the problem and the data to the machine.

• has led to increasing marriage between communications and computers.

• has in turn led to idea of computer utilities.

THE COMPUTER UTILITY

Built on the two ideas that

• there is an economy of scale

• it is feasible to move problems to machines and still affect an overall economy.

Other ideas:

• can give access to large system facilities for the time needed only.

• relieves the short staff problem for users.

But is this the only way to go?

Concurrent with the building of larger machines, smaller machines are becoming available.

• these machines have a place

• process control remote locations (including space)

• on-line research

But some parts of a computer are hard to reduce in size.

Consider the 3 main components:

• processor – can be reduced now

• storage – the current big push

• I/O Not too easy to do

There is a limit for the time being on how small a really adequate general purpose computer can get.

The ‘hand’ computer is not ever envisaged by Star Trek. Would need:

• voice input and output

• cartridge memories

• plug in for update

Real reason for the alternate approach is the latter, i.e. need for constantly up-to-date data banks.

Therefore better idea is to communicate as needed to a large machine.

How do you access a large computer remotely?

• high speed terminals

• low speed interactive terminals

Difference between time slicing approach and queuing for regular multiprogramming use.

Think of communication lines as a pipe.

High speed terminals may rent from $1,000 to $4,000 per month.

Low speed – $75 to $150, could be TWX or Teletype.

Latter type could give access to very small user but only with pre-packaged solution to his problems.

Can envisage a wide variety of specialized packages.

• real estate

• brokerage

Can we get smaller still?

Yes: • using voice answer over Touch Tone phone; • desk calculator;
• acoustic couplers; • remote sales reports; • grocery orders

Consider the mass referendum? Back to true Athenian democracy.

Considering this new point of view, let us generalize about application areas using remote access.

1. Financial Applications

Aimed at eliminating the flood of paperwork – the cashless society:

• central banking

• central stock registry

2 . Information Applications

The central data bank idea:

Could be factual only, e.g. const. data, library reference

Could involve manipulation

• models of economy

• DBS data

Languages developed for this – GIS, RPG

3. Computational Applications

Vary from straight mathematical to highly specialized – e.g. COGO, QUIKTRAN

A Problem

A basic problem of having many users on one large system is accounting for who uses what and how each should be charged. To date the lack of a generalized solution has slowed the development of a large multipurpose utility. A solution has only recently been proposed.

ADVANCED APPLICATIONS

To foresee where computers will be used next, examine the problem areas facing mankind.

1. Education – CAI

– learning from ETV

• a new approach to keeping up with a rapidly expanding base of information.

• we must teach basics and methods and let the computers retrieve the facts.

2. Transportation should we have to move ourselves to move our ideas?

• conferences over communication lines with mutual access to computerized information.

3. Urban Renewal – massive studies of where people live, move, work, etc. can lead to better urban development.

4. Population – vast computerized explanation of land resources

• CLI

Satellite controlled irrigation

Location of oceanographic food sources, e.g. schools of fish

5. Medicine Studies of medical histories can lead to cures by discovering trends or unusual relationships.

Sociological Implications

The real future for the computer is in solving the ‘people’ problems of the 20th century.

This use, however, has implications of which we should be aware.

• Changing role of govt. = no longer is it laissez-faire but is assuming an active planning role

• must have detailed data

• means a person’s private life becomes more of an open file

Where have we come?

• cashless society implies much of this already

• FBI National Crime Info. Centre

• Social Security

• Insurance Files

• Manpower Planning

etc. already have vast files.

Proposal in U.S. to establish 4 basic files in a National Databank

• Education (including test results)

Employment

Financial National Citizenship (including criminal files)

The whole idea of ‘keeping track of the individual’ could go further.

• already could check his every physical movement

• voice prints could check his every word.

The problem we must resolve is to obtain a balance between the absolute value of individual privacy and the absolute necessity of detailed data for social planning.

Some suggested safeguards:

1. Limit circulation of data

2. Review of data by individual

3. Review of who has requested data

4. Legal resource for misuse.

SUMMARY

Computers will ultimately be able to probe where man cannot go

• far reaches of space

â• depths of the earth

The gradual integration of man and machine is not beyond possibility.

Where we will see machines ultimately is limited only by our imagination.

Where we will see machines next depends on our point of view and I hope I have expanded this somewhat today.