Tag Archives: machine
#436184 Why People Demanded Privacy to Confide ...
This is part four of a six-part series on the history of natural language processing.
Between 1964 and 1966, Joseph Weizenbaum, a German American computer scientist at MIT’s artificial intelligence lab, developed the first-ever chatbot [PDF].
While there were already some rudimentary digital language generators in existence—programs that could spit out somewhat coherent lines of text—Weizenbaum’s program was the first designed explicitly for interactions with humans. The user could type in some statement or set of statements in their normal language, press enter, and receive a response from the machine. As Weizenbaum explained, his program made “certain kinds of natural-language conversation between man and computer possible.”
He named the program Eliza after Eliza Doolittle, the working-class hero of George Bernard Shaw’s Pygmalion who learns how to talk with an upper-class accent. The new Eliza was written for the 36-bit IBM 7094, an early transistorized mainframe computer, in a programming language that Weizenbaum developed called MAD-SLIP.
Because computer time was a valuable resource, Eliza could only be run via a time-sharing system; the user interacted with the program remotely via an electric typewriter and printer. When the user typed in a sentence and pressed enter, a message was sent to the mainframe computer. Eliza scanned the message for the presence of a keyword and used it in a new sentence to form a response that was sent back, printed out, and read by the user.
To encourage ongoing dialogue, Weizenbaum designed Eliza to simulate the type of conversational style used by a Rogerian psychoanalyst. The program would take something the user said and reflect it back with a question, as can be seen in the example below. (Notice how the program takes keywords like “boyfriend” and “depressed” and re-uses them.)
Human: Men are all alike.
Computer: IN WHAT WAY?
Human: They're always bugging us about something or other.
Computer: CAN YOU THINK OF A SPECIFIC EXAMPLE?
Human: Well, my boyfriend made me come here.
Computer: YOUR BOYFRIEND MADE YOU COME HERE
Human: He says I'm depressed much of the time.
Computer: I AM SORRY TO HEAR YOU ARE DEPRESSED
Human: It's true. I am unhappy.
Weizenbaum chose this mode of dialogue for Eliza because it gave the impression that the computer understood what was being said without having to offer anything new to the conversation. It created the illusion of comprehension and engagement in a mere 200 lines of code.
To test Eliza’s capacity to engage an interlocutor, Weizenbaum invited students and colleagues into his office and let them chat with the machine while he looked on. He noticed, with some concern, that during their brief interactions with Eliza, many users began forming emotional attachments to the algorithm. They would open up to the machine and confess problems they were facing in their lives and relationships.
During their brief interactions with Eliza, many users began forming emotional attachments to the algorithm.
Even more surprising was that this sense of intimacy persisted even after Weizenbaum described how the machine worked and explained that it didn’t really understand anything that was being said. Weizenbaum was most troubled when his secretary, who had watched him build the program from scratch over many months, insisted that he leave the room so she could talk to Eliza in private.
For Weizenbaum, this experiment with Eliza made him question an idea that Alan Turing had proposed in 1950 about machine intelligence. In his paper, entitled “Computing Machinery and Intelligence,” Turing suggested that if a computer could conduct a convincingly human conversation in text, one could assume it was intelligent—an idea that became the basis of the famous Turing Test.
But Eliza demonstrated that convincing communication between a human and a machine could take place even if comprehension only flowed from one side: The simulation of intelligence, rather than intelligence itself, was enough to fool people. Weizenbaum called this the Eliza effect, and believed it was a type of “delusional thinking” that humanity would collectively suffer from in the digital age. This insight was a profound shock for Weizenbaum, and one that came to define his intellectual trajectory over the next decade.
The simulation of intelligence, rather than intelligence itself, was enough to fool people.
In 1976, he published Computing Power and Human Reason: From Judgment to Calculation [PDF], which offered a long meditation on why people are willing to believe that a simple machine might be able to understand their complex human emotions.
In this book, he argues that the Eliza effect signifies a broader pathology afflicting “modern man.” In a world conquered by science, technology, and capitalism, people had grown accustomed to viewing themselves as isolated cogs in a large and uncaring machine. In such a diminished social world, Weizenbaum reasoned, people had grown so desperate for connection that they put aside their reason and judgment in order to believe that a program could care about their problems.
Weizenbaum spent the rest of his life developing this humanistic critique of artificial intelligence and digital technology. His mission was to remind people that their machines were not as smart as they were often said to be. And that even though it sometimes appeared as though they could talk, they were never really listening.
This is the fourth installment of a six-part series on the history of natural language processing. Last week’s post described Andrey Markov and Claude Shannon’s painstaking efforts to create statistical models of language for text generation. Come back next Monday for part five, “In 2016, Microsoft’s Racist Chatbot Revealed the Dangers of Conversation.”
You can also check out our prior series on the untold history of AI. Continue reading
Posted in Human Robots
#436176 We’re Making Progress in Explainable ...
Machine learning algorithms are starting to exceed human performance in many narrow and specific domains, such as image recognition and certain types of medical diagnoses. They’re also rapidly improving in more complex domains such as generating eerily human-like text. We increasingly rely on machine learning algorithms to make decisions on a wide range of topics, from what we collectively spend billions of hours watching to who gets the job.
But machine learning algorithms cannot explain the decisions they make.
How can we justify putting these systems in charge of decisions that affect people’s lives if we don’t understand how they’re arriving at those decisions?
This desire to get more than raw numbers from machine learning algorithms has led to a renewed focus on explainable AI: algorithms that can make a decision or take an action, and tell you the reasons behind it.
What Makes You Say That?
In some circumstances, you can see a road to explainable AI already. Take OpenAI’s GTP-2 model, or IBM’s Project Debater. Both of these generate text based on a large corpus of training data, and try to make it as relevant as possible to the prompt that’s given. If these models were also able to provide a quick run-down of the top few sources in that corpus of training data they were drawing information from, it may be easier to understand where the “argument” (or poetic essay about unicorns) was coming from.
This is similar to the approach Google is now looking at for its image classifiers. Many algorithms are more sensitive to textures and the relationship between adjacent pixels in an image, rather than recognizing objects by their outlines as humans do. This leads to strange results: some algorithms can happily identify a totally scrambled image of a polar bear, but not a polar bear silhouette.
Previous attempts to make image classifiers explainable relied on significance mapping. In this method, the algorithm would highlight the areas of the image that contributed the most statistical weight to making the decision. This is usually determined by changing groups of pixels in the image and seeing which contribute to the biggest change in the algorithm’s impression of what the image is. For example, if the algorithm is trying to recognize a stop sign, changing the background is unlikely to be as important as changing the sign.
Google’s new approach changes the way that its algorithm recognizes objects, by examining them at several different resolutions and searching for matches to different “sub-objects” within the main object. You or I might recognize an ambulance from its flashing lights, its tires, and its logo; we might zoom in on the basketball held by an NBA player to deduce their occupation, and so on. By linking the overall categorization of an image to these “concepts,” the algorithm can explain its decision: I categorized this as a cat because of its tail and whiskers.
Even in this experiment, though, the “psychology” of the algorithm in decision-making is counter-intuitive. For example, in the basketball case, the most important factor in making the decision was actually the player’s jerseys rather than the basketball.
Can You Explain What You Don’t Understand?
While it may seem trivial, the conflict here is a fundamental one in approaches to artificial intelligence. Namely, how far can you get with mere statistical associations between huge sets of data, and how much do you need to introduce abstract concepts for real intelligence to arise?
At one end of the spectrum, Good Old-Fashioned AI or GOFAI dreamed up machines that would be entirely based on symbolic logic. The machine would be hard-coded with the concept of a dog, a flower, cars, and so forth, alongside all of the symbolic “rules” which we internalize, allowing us to distinguish between dogs, flowers, and cars. (You can imagine a similar approach to a conversational AI would teach it words and strict grammatical structures from the top down, rather than “learning” languages from statistical associations between letters and words in training data, as GPT-2 broadly does.)
Such a system would be able to explain itself, because it would deal in high-level, human-understandable concepts. The equation is closer to: “ball” + “stitches” + “white” = “baseball”, rather than a set of millions of numbers linking various pathways together. There are elements of GOFAI in Google’s new approach to explaining its image recognition: the new algorithm can recognize objects based on the sub-objects they contain. To do this, it requires at least a rudimentary understanding of what those sub-objects look like, and the rules that link objects to sub-objects, such as “cats have whiskers.”
The issue, of course, is the—maybe impossible—labor-intensive task of defining all these symbolic concepts and every conceivable rule that could possibly link them together by hand. The difficulty of creating systems like this, which could handle the “combinatorial explosion” present in reality, helped to lead to the first AI winter.
Meanwhile, neural networks rely on training themselves on vast sets of data. Without the “labeling” of supervised learning, this process might bear no relation to any concepts a human could understand (and therefore be utterly inexplicable).
Somewhere between these two, hope explainable AI enthusiasts, is a happy medium that can crunch colossal amounts of data, giving us all of the benefits that recent, neural-network AI has bestowed, while showing its working in terms that humans can understand.
Image Credit: Image by Seanbatty from Pixabay Continue reading
Posted in Human Robots
#436155 This MIT Robot Wants to Use Your ...
MIT researchers have demonstrated a new kind of teleoperation system that allows a two-legged robot to “borrow” a human operator’s physical skills to move with greater agility. The system works a bit like those haptic suits from the Spielberg movie “Ready Player One.” But while the suits in the film were used to connect humans to their VR avatars, the MIT suit connects the operator to a real robot.
The robot is called Little HERMES, and it’s currently just a pair of little legs, about a third the size of an average adult. It can step and jump in place or walk a short distance while supported by a gantry. While that in itself is not very impressive, the researchers say their approach could help bring capable disaster robots closer to reality. They explain that, despite recent advances, building fully autonomous robots with motor and decision-making skills comparable to those of humans remains a challenge. That’s where a more advanced teleoperation system could help.
The researchers, João Ramos, now an assistant professor at the University of Illinois at Urbana-Champaign, and Sangbae Kim, director of MIT’s Biomimetic Robotics Lab, describe the project in this week’s issue of Science Robotics. In the paper, they argue that existing teleoperation systems often can’t effectively match the operator’s motions to that of a robot. In addition, conventional systems provide no physical feedback to the human teleoperator about what the robot is doing. Their new approach addresses these two limitations, and to see how it would work in practice, they built Little HERMES.
Image: Science Robotics
The main components of MIT’s bipedal robot Little HERMES: (A) Custom actuators designed to withstand impact and capable of producing high torque. (B) Lightweight limbs with low inertia and fast leg swing. (C) Impact-robust and lightweight foot sensors with three-axis contact force sensor. (D) Ruggedized IMU to estimates the robot’s torso posture, angular rate, and linear acceleration. (E) Real-time computer sbRIO 9606 from National Instruments for robot control. (F) Two three-cell lithium-polymer batteries in series. (G) Rigid and lightweight frame to minimize the robot mass.
Early this year, the MIT researchers wrote an in-depth article for IEEE Spectrum about the project, which includes Little HERMES and also its big brother, HERMES (for Highly Efficient Robotic Mechanisms and Electromechanical System). In that article, they describe the two main components of the system:
[…] We are building a telerobotic system that has two parts: a humanoid capable of nimble, dynamic behaviors, and a new kind of two-way human-machine interface that sends your motions to the robot and the robot’s motions to you. So if the robot steps on debris and starts to lose its balance, the operator feels the same instability and instinctively reacts to avoid falling. We then capture that physical response and send it back to the robot, which helps it avoid falling, too. Through this human-robot link, the robot can harness the operator’s innate motor skills and split-second reflexes to keep its footing.
You could say we’re putting a human brain inside the machine.
Image: Science Robotics
The human-machine interface built by the MIT researchers for controlling Little HERMES is different from conventional ones in that it relies on the operator’s reflexes to improve the robot’s stability. The researchers call it the balance-feedback interface, or BFI. The main modules of the BFI include: (A) Custom interface attachments for torso and feet designed to capture human motion data at high speed (1 kHz). (B) Two underactuated modules to track the position and orientation of the torso and apply forces to the operator. (C) Each actuation module has three DoFs, one of which is a push/pull rod actuated by a DC brushless motor. (D) A series of linkages with passive joints connected to the operator’s feet and track their spatial translation. (E) Real-time controller cRIO 9082 from National Instruments to close the BFI control loop. (F) Force plate to estimated the operator’s center of pressure position and measure the shear and normal components of the operator’s net contact force.
Here’s more footage of the experiments, showing Little HERMES stepping and jumping in place, walking a few steps forward and backward, and balancing. Watch until the end to see a compilation of unsuccessful stepping experiments. Poor Little HERMES!
In the new Science Robotics paper, the MIT researchers explain how they solved one of the key challenges in making their teleoperation system effective:
The challenge of this strategy lies in properly mapping human body motion to the machine while simultaneously informing the operator how closely the robot is reproducing the movement. Therefore, we propose a solution for this bilateral feedback policy to control a bipedal robot to take steps, jump, and walk in synchrony with a human operator. Such dynamic synchronization was achieved by (i) scaling the core components of human locomotion data to robot proportions in real time and (ii) applying feedback forces to the operator that are proportional to the relative velocity between human and robot.
Little HERMES is now taking its first steps, quite literally, but the researchers say they hope to use robotic legs with similar design as part of a more advanced humanoid. One possibility they’ve envisioned is a fast-moving quadruped robot that could run through various kinds of terrain and then transform into a bipedal robot that would use its hands to perform dexterous manipulations. This could involve merging some of the robots the MIT researchers have built in their lab, possibly creating hybrids between Cheetah and HERMES, or Mini Cheetah and Little HERMES. We can’t wait to see what the resulting robots will look like.
[ Science Robotics ] Continue reading
Posted in Human Robots