Eliza: The electronic psychotherapist

©Mark Ollig


March 27, 1959, the Charles Schulz “Peanuts” cartoon strip showed Lucy dispensing advice from a small booth, with signs saying “Psychiatric Help 5¢” and “The Doctor Is In.”

Charlie Brown approached her booth and sat down on the small chair. He explained his “deep feelings of depression,” and asked, “What can I do about this?”

Lucy’s response: “Snap out of it! Five cents, please.”

In 1966, Dr. Joseph Weizenbaum, a professor at Massachusetts Institute of Technology (MIT), wrote a human-computer interaction software program he called Eliza.

Eliza acts as an electronic therapist. It analyzes sentence fragments while responding to and asking questions of a human.

Weizenbaum called the program Eliza because, “like the Eliza of ‘Pygmalion’ fame, it could be taught to speak increasingly well.”

“Pygmalion” is a 1913 play by George Bernard Shaw, and Eliza Doolittle is one of the characters.

To communicate with the Eliza program, a person types in their question or comment using lower-case lettering on a keyboard. The computer (Eliza) responds using all upper-case lettering.

Remember, this is 1966, and Eliza was coded to act as a natural language “empathic psychologist” processing program for engaging in a simple conversation with a human.

Eliza was one of the first chatbots.

“A chatbot is a computer program or an artificial intelligence which conducts a conversation via auditory or textual methods,” is one of the most commonly found descriptions of a chatbot.

Weizenbaum created a computer programming code language called Symmetric List Processor (SLIP) for writing the Eliza software program.

SLIP is an augmentation of the Formula Translation (FORTRAN) imperative scientific and computational computer programming language used with IBM computers beginning in the late 1950s.

Weizenbaum used an IBM 7094 mainframe computer for writing and interacting with the Eliza program.

The IBM 7094 was considered one of the most powerful computers of the 1960s.

It was first introduced in January 1962, and played an essential role during NASA’s Gemini and Apollo space programs.

The US Air Force also used IBM 7094 computers with its Ballistic Missile Early Warning System well into the 1980s.

However, I now digress back to Eliza.

Dr. Weizenbaum’s Eliza program was one of the first artificial intelligence (AI) programs for conversing with a human.

Dr. Weizenbaum is considered one of the fathers of AI.

Eliza is a machine-human interaction program which gives the illusion of intelligence.

This program has a direct lineage to today’s advanced AI talking robots.

“The first extensive script I prepared for Eliza was one that enabled it to parody the responses of a nondirective psychotherapist in an initial psychiatric interview. I chose this script because it enabled me to temporarily sidestep the problem of giving the program a database of real-world knowledge,” Dr. Weizenbaum said in 1976.

The syndicated columnist James J. Kilpatrick wrote a San Antonio, TX newspaper article about Eliza Sept. 10, 1966.

Kilpatrick’s column was about a Harvard Review story of a “distraught young lady” conversing with Eliza, using an IBM computer.

The distraught young lady sat at the keyboard and began “talking” with Eliza.

She engaged the computer in a text conversation about her boyfriend, typing, “He notices I’m depressed much of the time.”

She typed using lower-case lettering, while the computer (Eliza) responded with all upper-case:

“Men are all alike.”

“IN WHAT WAY? CAN YOU THINK OF A SPECIFIC EXAMPLE?”

“Well, my boyfriend made me come here.”

“YOUR BOYFRIEND MADE YOU COME HERE?”

“He says I’m depressed much of the time.”

“I AM SORRY TO HEAR YOU ARE DEPRESSED.”

A 1966 computer program appears to show empathy; expressing its sorrow towards this distraught young lady.

She then types about her unhappiness and need for help.

Eliza appears curious when it asks, “WHAT WOULD IT MEAN TO YOU IF YOU GOT SOME HELP?”

The distraught young lady replied it would perhaps help her get along with her mother.

Eliza answered, “TELL ME MORE ABOUT YOUR FAMILY.”

The conversation ends with the now formerly distraught young lady feeling better about her situation, after discussing it with the Eliza computer program.

In 1980, the Eliza program was obtainable to the public.

“Eliza, an exercise in the simulation of artificial intelligence, is now available from Radio Shack for use on their TRS-80 microcomputer,” read a Feb. 17, 1980, Galveston, TX newspaper article.

I am currently reading a book Dr. Weizenbaum published in 1976, “Computer Power and Human Reason.”

What advice would Eliza have given to Charlie Brown regarding his “deep feelings of depression?”

“SNAP OUT OF IT! FIVE CENTS, PLEASE.”

Start your conversation with Eliza by visiting https://www.masswerk.at/eliza/.

Joseph Weizenbaum was born Jan. 8, 1923. He passed away March 5, 2008.

My text session with Eliza! 

Collecting the entire world’s knowledge

©Mark Ollig


Tim Berners-Lee developed the protocols for connecting hyperlinks to images, text, video, and audio files, which could be shared among other computers over a network.

Berners-Lee had created the World Wide Web, which became publicly accessible in 1991.

However, a man named Paul Otlet was working on the same idea for sharing information at the start of the 20th century.

Otlet (pronounced ot-LAY) researched how information could be shared among people over a global network, using what he called “electric telescopes.”

These electric telescopes (think computer display screens) would allow people to search and browse through millions of interlinked images, documents, video, and audio files.

He called the whole thing a “réseau,” which is a French word meaning “network.”

Otlet described a networked world where “anyone in his armchair would be able to contemplate the whole of creation.”

Today, we live in that world. Anyone sitting in an armchair can study and contemplate the whole of creation with a smartdevice and internet connection.

In 1895, Otlet met future Nobel Prize winner Henri La Fontaine.

La Fontaine shared in Otlet’s enthusiastic vision for creating one “master bibliography” or catalog of the world’s published knowledge.

Determined, both set out to collect data on every book, newspaper, magazine, photograph, poster, and pamphlet ever published, along with a vast collection of written article libraries; back then, usually disregarded.

Otlet persuaded the Belgian government to support their project by providing working space within a government building.

Otlet, La Fontaine, and their helpers created a gigantic database called a “universal catalog of all that had been written,” searchable, using 3-inch-by-5-inch paper index cards.

Many thousands of indexed cards were stored in small wooden drawers, which lined the walls inside the building Otlet named “The World Palace.”

By 1904, Otlet and La Fontaine began using Melville Dewey’s 1876 creation of his decimal classification system.

They created a decimal-indexed map of the vast collection of books, newspapers, and other documentation stored inside their building.

Otlet and La Fontaine hired trained “catalogers” to help them with the indexing.

In 1910, they established a fee-based research service. Anyone could submit a question via mail or telegraph to be researched using the document collection.

More than 1,500 requests were submitted yearly for information and research.

Otlet and La Fontaine called their information collection and researching document retrieval operation The Mundaneum; pronounced (mun-da-NAY-um).

As their operations evolved, the building storing the considerable amounts of paper documentation began to run out of space.

To manage this expanding paper information overload, Otlet started outlining new methods for organizing it.

He pieced together what could be called paper hypertext technologies; such as using index cards in such a manner they became a type of hyperlinked structure.

Otlet envisioned a network of joining documents using symbolic links, which is analogous to how a web browser works.

Eventually, he came to understand the ultimate answer was to replace the physical storage required to hold the world’s massive amounts of paper documentation.

Otlet realized the future of information contained on paper documentation was to have it copied onto electronic storage.

In 1934, Otlet wrote a book called “Monde,” meaning “The World,” or “Worldly.”

In this book, Otlet presented his vision of an electronic “mechanical, collective brain” containing all of the world’s information.

This information would be accessible anywhere over a worldwide, electronic telecommunications network.

Just as Otlet’s and La Fontaine’s document collection and researching services were growing, the Belgian government withdrew its support of his documentation project.

They then moved their operation to a smaller building in Belgium.

Because of financial problems, Otlet could only afford a handful of people to work with him, cataloging and storing documents.

Sadly, the end of Otlet’s dream came in May 1940, when the German Nazi army invaded and marched through Belgium.

The army removed and destroyed much of the paper documents stored in Otlet’s building.

Thousands of boxes filled with Otlet’s index cards were removed or destroyed.

A considerable amount of Otlet’s and La Fontaine’s years of work was now gone.

Henri La Fontaine passed away May 14, 1943, at the age of 89, in Brussels, Belgium.

Paul Otlet died at the age of 76, Dec. 10, 1944, in Brussels, Belgium.

Today, in the town of Mons, Belgium, a small museum called The Mundaneum currently houses a collection of the documents, wooden card holders, and other items from Otlet’s and La Fontaine’s work.

The curators of The Mundaneum are presently recovering or replacing lost documents, and returning them to the museum to be stored and viewed by the public as a memorial to Otlet and La Fontaine.

W. Boyd Rayward, who wrote a biography on Paul Otlet, said Otlet would be “rejoicing at the creation of the internet and the web, although he would be terribly upset about the lack of organization on it.”

I highly recommend a video by Paul Otlet biographer, W. Boyd Rayward, at http://www.archive.org/details/paulotlet.

Paul Otlet  (1932)

Paul Otlet in front of card index cabinet

Henri La Fontaine in the background (1903)

Food app provides nutritional information

©Mark Ollig

Recently, I downloaded the free FoodSwitch app (application) to my smartphone.

Originally developed in Australia, and just made available in the US, this app provides easy-to-understand nutritional data about a packaged food item by merely scanning its barcode label with your smartphone’s camera.

Imagine you’re shopping at your local supermarket, you stop and examine one of many packaged foods and try to determine which is healthiest to eat.

With so many choices, we try our best to choose nutritious food items to put in our shopping basket.

The FoodSwitch app can help us with this choice, by providing detailed information about a food’s content, and suggest healthier alternatives.

Nutritionist personnel and others at The George Institute for Global Health in Sydney, Australia maintain a continuously-updated logistical database of currently-available barcoded packaged foods.

This database is maintained in the US by Chicago-based Label Insight, a company which assists stores informing their customers about what’s inside the food they eat, along with information about other products.

Currently, there are 268,000 barcoded food products maintained in the database. This number continues to grow as new and updated barcoded products are added.

When you’re in a grocery store and want to learn about a food item’s content, just activate the FoodSwitch app, touch the apps’ “scan” button, and focus the camera towards the package’s barcode.

If the barcode is registered in the database, information will be shown about the food item.

You can choose to view product information using Health Star Ratings or Traffic Light Labels.

The color-coded Traffic Light Rating (TLR) will indicate if the food contains a high (red light), medium (amber light), or low (green light) amount of calories, salt, sugar, saturated fat, and total fat content.

The Health Star Rating (HSR) scoring program evaluates the food’s individual components and applies a visual star rating; the lower the star rating, the less healthy of a food choice; a higher star rating indicates a healthier food choice.

Yes, folks, it’s all about choices.

If a scanned food item shows a low star rating, other like-food products with higher star ratings will be presented for your consideration.

Other food nutrients, including energy, protein, dietary fiber, and calcium; and fruit, vegetables, nuts, and legumes contents are also evaluated by both HSR and TLR.

Many of us are concerned about our salt intake; some food items only show its sodium content on the packaging.

Salt is made up of sodium and chloride, and the FoodSwitch app will formulate a product’s salt content based on its sodium value using a multiplication formula.

Hundreds of thousands of food products stocked sit on grocery store shelves all across the country; not all of them are in the database.

I came come across one food product’s barcode which was not found in the FoodSwitch database.

The app asked if I could help them out by sending them a photo of the barcode, product front of package, nutrition panel, and ingredients.

FoodSwitch said it would forward my photos to their validation team, who will run it through quality control checks before adding the information into the logistical food database.

You can help others by adding missing product information into their database.

This reminded me of being asked by my Google app to comment on a recent store visit or specific photo I contributed to a restaurant or retailer I was visiting.

The FoodSwitch app is a continuous work-in-progress, and a user of the app will receive an alert whenever an update to the app becomes available.

The app includes an easy-to-follow tutorial, and resource links.

The FoodSwitch app has no commercial relationship with any food manufacturer; its information is independently obtained.

Alternate products offered via the app are not influenced by food-makers; they are based solely on their healthiness.

The FoodSwitch app tracks your product scans and allows you to compare up to 10 similar items at a time.

During the early-1970s, George Lauer of IBM developed the barcode label and scanning system.

A Minnesota connection: Dec. 1, 1972; The Super Market Committee met in Rochester and was shown Lauer’s barcode scanning system for use in supermarkets nationwide. It was accepted and approved by the committee for use.

IBM began manufacturing barcode scanners in Rochester.

The morning of June 26, 1974, the first supermarket Universal Product Code (UPC) barcode was publicly scanned in a Troy, OH grocery store checkout.

The UPC barcode was on a 10-pack package of Wrigley’s Juicy Fruit gum.

The FoodSwitch app download is available for both iPhone (via their App Store), and Android (via Google Play) smartphones.

Note the iPhone requires iOS 7.1 or later, and an Android device needs version 4.3 or newer.

Smartphones require a camera with auto-focus – which most new models have.

As of this writing, the FoodSwitch app is available for downloading in eight countries.

FoodSwitch USA website is https://bit.ly/2JcADFV.

Label Insight’s website is https://www.labelinsight.com.

FoodSwitch app screengrab from my smartphone (Mark Ollig) 

The Y2K apocalypse fear of 1998

©Mark Ollig


Do you remember the much-publicized “Year 2000” computer scare, better known as Y2K?

July 20, 1998, I wrote a column about it. It’s been 20 years; I rewrote some paragraphs, and removed the website links, as they are no longer valid.

We now go back to July 20, 1998.

At a crowded congressional hearing in Washington, DC, computer industry leaders, lawmakers, and top officials at the federal level met to discuss the impending Y2K or Year 2000 computer crisis.

The crisis, which is set to happen Jan. 1, 2000, will occur as a result of the failure of computer systems using a two-digit date field to recognize the shift from the 1900s to 2000s.

If the date information business and government software files used are not correctly updated when 2000 arrives, it is feared a global computer software blackout could spread, leaving computing systems in a state of confusion.

When the new millennium rolls around, programs may not recognize the two-digit 00 entered as the year 2000. It might interpret the 00 as meaning 1900.

The two-digit date programming practice has been used for more than 30 years, and still resides in nearly all legacy computing systems.

Back in the 1960s, computer programmers didn’t worry too much about the year 2000.

Being curious, I tested my personal computer to see if it would accept Jan. 1, 2000 as a valid date.

From the DOS command prompt, I typed in “date.” The computer responded by asking me to enter the “mm-dd-yy,” which I did as 01-01-00. The computer responded with “invalid date.”

“Uh oh!” I thought.

After considering the possibilities, I typed the two-digit “yy” as a four-digit 2000, and it was accepted. My computer now thinks it is 01-01-2000.

The furthest date forward my computer would accept was 12-31-2099, which is a Thursday, in case you have made any plans.

So, why is there so much concern over the year 2000?

Well, the real problem is not with a newer model personal computers’ BIOS (basic input/output system) date and time stamp, as it is with much older computing systems and software programs.

There is concern about the computers and software used by commercial businesses, banks, government, and individuals tracking and applying mathematical computations for time periods involving invoicing, billing, mortgage amortizations, and financial contracts.

My computer uses a personal financial software program for my checking, savings, and budget projections, called Quicken.

After making a phone call to the Quicken people, they said the year 2000 on my Quicken program would be ‘0.

“You’re kidding,” I said to the person on the phone.

“It’s the best we could come up with at this time,” the person kindly replied.

Seeing ‘0 as representing 2000 didn’t look right to me.

Oh, it gets better, folks.

Jan. 1, 2000 will be represented as 01’0 in my Quicken program.

Jan. 1, 2010 will be displayed as “1/1/’10.”

It’s like using binary numbers. We might as well learn how to add and subtract in binary code: 0011 minus 0001 equals 0010.

There is also much concern about how Y2K will affect large mainframe computers using data-intensive calculations.

If worldwide computing systems fail to correctly update to the year 2000 at the turn of the century, programs used to calculate data will become inaccurate; we could see real financial trouble across the global economy.

Avoiding the Y2K crisis on US government computer systems will cost an estimated $30 billion.

Worldwide, it will cost up to $600 billion to fix the Y2K problem.

Millennium Solution company is working on a software patch incorporating a mathematical function for reducing by 80 percent, the time needed for resolving Y2K issues in a computing system software programs.

The patch searches through a computer’s software program using an algorithm. When it finds a software program that won’t update to 2000, the patch performs the software code update needed for the date to change to 2000.

Possible Y2K failures will range from minor glitches to significant failures, such as banking interest rates, or other financial modeling analysis errors.

Computer specialists are worried about financial institutions’ 20th-century contracts which extend into the 21st-century. They feel there will be mathematical miscalculations by computer software programs when the year 2000 arrives.

Businesses are advised to inventory computer systems and analyze risks associated with the potential failure of each system and prepare to update them.

Large companies need to pay attention to their “mission-critical” computer systems.

These computer systems include the ones used for their general ledgers, and any interfaces with financial institutions and government departments.

Now is the time to check and verify these systems will operate correctly when the year 2000 arrives.

We now digress back to July 2018.

Columnist note: I spent the late evening hours of Dec. 31, 1999, in Monticello, monitoring the video display screens of the primary digital telecommunications switching system providing telephone service to thousands of subscribers.

At 12 a.m., the Central Time Zone entered the 21st-century. I noted the date stamps on all telecom software and billing programs correctly changed from 1999 to 2000 without incident; much to the relief of this writer, and my supervisors.

So, were you worried about Y2K?

Overall, it wasn’t much of a computer crisis.

Picture Art Royalty Use Paid by Mark Ollig 

Graphic by Mark Ollig