Galáctica y las dificultades de los modelos de lenguaje

En noviembre, Meta presentó un modelo de lenguaje bautizado Galactica, elaborado para asistir a investigadores científicos, pero sólo tres días después fue retirado de disponibilidad para ser consultado o testeado. Básicamente, como ha sucedido en otros campos de trabajo con inteligencia artificial (IA/AI), el lenguaje no reconoce verdad o falsedad. En las pruebas, trabajos formalmente presentados como científicos pero absurdos como la existencia de osos en el espacio, o las causas de la guerra de Ucrania, pasaron por buenos, con justificaciones razonadas.

Will Douglas Heaven, en Technology Review:

Galactica is a large language model for science, trained on 48 million examples of scientific articles, websites, textbooks, lecture notes, and encyclopedias. Meta promoted its model as a shortcut for researchers and students. In the company’s words, Galactica “can summarize academic papers, solve math problems, generate Wiki articles, write scientific code, annotate molecules and proteins, and more.”

(...) A fundamental problem with Galactica is that it is not able to distinguish truth from falsehood, a basic requirement for a language model designed to generate scientific text. People found that it made up fake papers (sometimes attributing them to real authors), and generated wiki articles about the history of bears in space as readily as ones about protein complexes and the speed of light. It’s easy to spot fiction when it involves space bears, but harder with a subject users may not know much about.

(...) Many scientists pushed back hard. Michael Black, director at the Max Planck Institute for Intelligent Systems in Germany, who works on deep learning, tweeted: “In all cases, it was wrong or biased but sounded right and authoritative. I think it’s dangerous.”

(...) The Meta team behind Galactica argues that language models are better than search engines. “We believe this will be the next interface for how humans access scientific knowledge,” the researchers write.  This is because language models can “potentially store, combine, and reason about” information. But that “potentially” is crucial. It’s a coded admission that language models cannot yet do all these things. And they may never be able to. “Language models are not really knowledgeable beyond their ability to capture patterns of strings of words and spit them out in a probabilistic manner,” says [Chirag Shah,  University of Washington]. “It gives a false sense of intelligence.”

 Grady Booch comenta: "Galactica is little more than statistical nonsense at scale. Amusing. Dangerous. And IMHO unethical". Algún investigador en ML (Yann LeCun, en el mismo hilo), se escandaliza por la calificación de no ético. Creo que a algunos científicos les falta medir el alcance de lo que tienen entre manos.

 

 

Frederick Brooks: muere un pionero

 

Hace pocos días, el 17 de noviembre, ha muerto Frederick Brooks, un pionero de la ingeniería de software, casi de su primera generación. Longevo, continuó trabajando vinculado a las tecnologías digitales hasta la primera década de este siglo, comenzando desde 1953, después de egresar de la Universidad de Duke. Pasó por IBM a partir de 1956 y hasta 1965, donde dirigió el diseño de los ordenadores 360 (IBM System/360), el primer mainframe de IBM, base de la arquitectura estructurada por IBM, y padre directo de los 4300 y los actuales System Z. Todavía hoy una aplicación codificada en y para el 360 puede ejecutarse en un System/Z. En las decisiones que permitieron esta evolución, uno de los pilares fue Brooks. 

El otro gran aporte de Brooks está en la metodología, en la sistematización de su experiencia en sus años de IBM, en primer lugar, en 1975, con The Mythical Man-Month, y años después, en 1986, con No Silver Bullet—Essence and Accident in Software Engineering, agregado luego como nuevo capítulo en The Mythical...Existe un gran salto entre las épocas en que escribió estos libros y su lectura actual, pero a pesar del desfase técnico, todavía deben ser libros de lectura obligatoria. 

Lo que sigue es el obituario de Shane Hastie en InfoQ, con un buen conjunto de referencias a los logros de Brooks:

Dr Frederick P Brooks Jr, originator of the term architecture in computing, author of one of the first books to examine the nature of computer programming from a sociotechnical perspective, architect of the IBM 360 series of computers, university professor and person responsible for the 8-bit byte died on 17 November at his home in Chapel Hill, N.C. Dr Brooks was 91 years old.

He was a pioneer of computer architecture, highly influential through his practical work and publications including The Mythical Man Month, The Design of Design and his paper No Silver Bullet which debunked many of the myths of software engineering.

In 1999 he was awarded a Turing Award for landmark contributions to computer architecture, operating systems, and software engineering. In the award overview it is pointed out that

Brooks coined the term computer architecture to mean the structure and behavior of computer processors and associated devices, as separate from the details of any particular hardware implementation

In the No Silver Bullet article he states:

There is no single development, in either technology or management technique, which by itself promises even one order-of-magnitude improvement within a decade in productivity, in reliability, in simplicity.

Quotations from the Mythical Man Month:Essays on Software Engineering permeate software engineering today, including:

• Adding manpower to a late software project makes it later.  
• The bearing of a child takes nine months, no matter how many women are assigned.
• All programmers are optimists.

On April 29, 2010 Dilbert explored the adding manpower quote.  

In 2010 he was interviewed by Wired magazine. When asked about his greatest technical achievement he responded

The most important single decision I ever made was to change the IBM 360 series from a 6-bit byte to an 8-bit byte, thereby enabling the use of lowercase letters. That change propagated everywhere.

He was the founder of the Computer Science Department at the University of North Carolina at Chapel Hill, where the Computer Science building is named after him. In an obituary the University says:

Dr. Brooks has left an unmistakable mark on the computer science department and on his profession; this is physically recognized by the south portion of the department’s building complex bearing his name. He set an example of excellence in both scholarship and teaching, with a constant focus on the people of the department, treating everyone with respect and appreciation. His legacy will live on at UNC-Chapel Hill

His page on the university website lists his honours, books and publications.

The Computer History Museum has an interview of Dr Brooks by Grady Booch.

He leaves his wife of 66 years Nancy, three children, nine grandchildren and two great-grandchildren.