The philosophy of the lab was simple: there’s a ton of resources here, physical and intellectual, so go do interesting stuff. I soon teamed up informally with Bruce Blumberg, whose focus was on learning how dogs make decisions, and in building computer systems that learn the kinds of things that animals learn easily. He also had a dog, a silky terrier, a cute little thing that Bruce said was his inspiration. Our intellectual interests overlapped a lot, and because he was the “dog guy” and I was the “bird lady,” we called ourselves the Woofers and Tweeters.
Before long, I was working on various projects with some of Bruce’s graduate students, Ben Resner in particular. One project was Serial Tr-Hacking (“hacks” being MIT practical jokes based on technology), in which Wart had to identify instructions, in the form of simple images, and then perform a combination of pulling, flipping, and rotating a lever to have a food treat delivered to him. Another was an electronic bird sitter, designed to stop the screeching parrot problem. This involved a screen on which the parrot watched pictures and videos. Exactly what images were displayed depended on how much noise the bird was making. Below a desired level, the pictures would be positive images, such as the bird’s owner, parrots in the wild, that kind of thing. If the parrot’s whistles and screeches exceeded the desired level, the pictures would switch to negative objects, such as a raptor swooping or a ground predator creeping nearby. The idea was that a microphone system would track the sound level of the bird’s calls and control the images accordingly.
We were also thinking about something we called a “smart nest,” a device to track Greys’ behavior at nest sites in Africa. We were working on global positioning system tags that were small enough and light enough to be attached to Greys’ backs, so we would be able to monitor their movements during the day.
And we were developing a system that might be used both for enriching a parrot’s vocabulary and for working with autistic children. We imagined having a series of “toys,” each of which would have a radio frequency identification tag. The action of picking up a toy would trigger the playing of a video about the toy. For instance, if the toy was a key (parrots used keys for scratching themselves), then the video would be of a person saying something like, “A key. You’ve got a key. Wow! Look at the key.” When the toy was put down, the video would stop. That was the idea, anyway. We called it our PollyGlot Computer. The challenge was to have enough variation in the videos so that the subject wouldn’t get bored.
If that one sounds a little sterile, it was. Boredom was a constant issue with Wart. It was hard to make these games interesting enough to keep his attention for very long. If an undergraduate entered the room during a session, Wart was likely to find the newcomer far more interesting than pulling or pushing a lever one more time. Ben used to say that he and his fellow students felt they were in competition with Wart: the students were trying to outsmart Wart, while Wart was busy outsmarting the students. The students told me they felt as if Wart were saying, Hey, give me a real problem. This one is just too easy for a smart bird like me!
Bruce and I put on a mini-symposium, called Wired Kingdom, the spring after I arrived. We had people describe all kinds of ways of using electronic gadgets to study animals in the wild and in zoos. I was having a blast, not least because of the opportunity to interact with such bright and motivated students. I loved that. Apparently the Lab’s directors did, too, because shortly after our symposium they asked me to stay a second year, which was unheard of for a visiting professor, as I was. This time I was determined not to leave Alex in Tucson. He and Griffin were coming to Boston to be with me and Wart.
Wart was also having a blast. When I think about the birds’ personalities, I always come to an amusing contrast. Alex was like the kid in class who always knows the answers and is constantly jiggling around in his seat, his hand waving high, wanting to be the one to be chosen to answer the teacher. Griffin is like the smart but shy kid, trying to make himself invisible so he won’t be chosen. Wart is the party animal, the teenager playing hooky with his friends. He was also like the techno-freak teenager, perfect for his role at the Lab. He was good at manipulating the equipment, and he loved it. He was confident and at ease performing in front of an audience.
There were always audiences at the Media Lab. The Lab was funded by sponsors, corporations who got to see and benefit from what was going on in the Lab, for a sizeable fee. There were always visitors coming through, but the main events were two sponsor weeks, also known as “demo weeks,” one in the spring and the other in the fall, very lavish affairs. During the time running up to demo week, the Lab was a frenzy of activity as the students put the final tweaks on their demos, practically living there. The Lab’s saying, “Demo or die,” was taken very seriously.
Those were intoxicating times to be at the Lab. The stock market was setting records every day. Dot-com mania was in full bloom. The Lab seemed awash in corporate money. When I learned early on that my NSF grant had not been renewed, the Lab essentially said, “Don’t worry; just do your work.” I thought, Wow! This is not Tucson.
There was never any shortage of cool things for sponsors to see during sponsor week, but a live parrot in a demo was always an added attraction. Wart performed wonderfully at our first such event, in the spring of 2000. Every fifteen minutes or so a group of people would come through my lab, and we would demonstrate our projects. Wart would pull and flip levers, as instructed. He was a natural.
By the end of the week, the poor bird was exhausted. It was the last day, and one more sponsor came through. Wart was sitting absolutely still on his perch, eyes closed, napping. The sponsor came into the lab, saw Wart, and stooped to take a closer look. Wart slowly opened an eye, then closed it. Otherwise he was motionless. The sponsor exclaimed, “Oh! Animatronics!” I said, “No, no. He’s not a robot. He’s a real bird. Don’t get too close; you might get bitten!”
I was thrilled by the prospect of having Alex and Griffin with me again, and I looked forward to introducing them to this new world of technology. But there were hassles in getting them to Boston. I was finally able to bring them with me in separate carriers on a red-eye flight, connecting through Dallas. The poor birds were miserable from the time I put them in the carriers to when I let them out in Boston, some twelve hours later. They had refused to eat anything. I had taken them to the plane’s bathroom with me, trying to coax them with treats in private. But they were too stressed. Alex was a particularly sorry sight, his tail feathers all but chewed off. It is common for parrots to overpreen their feathers when they are distressed, and the months of our separation had been hard on him. I had seen that on my periodic visits to Tucson. Each time I got back he struggled with the pleasure of seeing me again and the anger at my being away from him. But now we were together.
Shortly after Alex and Griffin arrived at the Media Lab, in September 2000, a producer for Scientific American Frontiers, a public television science show, contacted me about taping a segment of an episode. I had done one a decade earlier, in Tucson. The current show was to be on pets and technology, called “Pet Tech.” Alan Alda was the presenter. I was going to meet Hawkeye Pierce!
When Rebecca, my unofficial godchild, heard about this, she was wild with excitement. She is a huge fan of M*A*S*H, and she begged me to get an autograph for her. I confess I was a little bit nervous before meeting Alda, unusual for me. But he also turned out to be a really nice guy, very amusing and friendly. When I proffered Rebecca’s copy of the M*A*S*H book and asked if he would autograph it for my godchild, he raised an eyebrow, just the way I’d seen on screen. “Of course,” he said. “And what is your…er…godchild’s name?” I told him it was Rebecca. He laughed and then said, “So it really is for your godchild!” I frowned, puzzled by what he meant. “Well, people often say it’s for their godchild, but it’s really for themselves, and they are too embarrassed to admit it.” He signed the book, and I gave him a copy of The Alex Studies in return.
It certainly was one of the m
ost fun taping sessions I had done, partly because of who Alda is and how very charming he is, and also because he was so obviously enthralled with Alex and what he could do. The show started with a few clips from the previous show, in which Alex identified the color of an object, responded to “How many?” and “What color bigger?” The shot then cut to Alda, me, and Alex in the Lab. “Hello, Alex,” he said. Then he turned to me and asked, “Can Alex do any new things since last time?”
I said, “Yes, he can,” and proceeded to demonstrate.
I showed Alex two keys and asked, “What toy?” I then asked him, “How many?” and “What’s different?” Alex was in good form and answered quickly and correctly, if a little mumbly on the final answer.
I then held up a tray of plastic Arabic numbers, each a different color. By this time, Alex had learned numbers up to six. “What number green?” I said.
Alex hesitated and said, “Want a nut.”
I said, “Come on, Alex, you can have a nut later. What number is green?” I was thinking, Oh, no. Is he going down that path?
But then he quickly said, “Four,” which was correct. Then he again said, “Want a nut.” I gave him one.
Alda was sitting there shaking his head, grinning widely, clearly impressed by Alex’s talents. I explained the model/rival technique for teaching Alex labels and concepts, and suggested we do a session right then and there. I asked Alda, “What toy?” holding up a spoon. We went back and forth on that.
Then Alda turned to Alex and said, “Alex, what toy?”
“Want a nut,” said Alex.
Eventually Alex produced a “sss” sound, and then something that approximated the full word. It is a difficult word for him to say.
“I find this very hard to believe,” said Alda, “that Alex is really doing what he seems to be doing.” He then turned to the camera and said that he was well used to coming to the Media Lab to film interesting projects. “I have to say that Alex and the other birds look oddly out of place in an institution best known for high technology.” He explained my pursuit of gadgets for entertaining parrots. “To find out why parrots need entertaining, we are going to go to Foster Parrots, about an hour south of Boston.”
Foster Parrots is a parrot rescue facility run by Marc Johnson. There we had a conversation about the problems people have with pet parrots that are left alone all day, the kind of thing I had been proselytizing about for years. “It is like putting a four-year-old kid in a playpen in the morning and leaving him there all day by himself,” I said. “Of course he would be angry and upset when you got home. It’s the same with parrots.”
“You have to remember that dogs have been domesticated for thousand of years,” explained Johnson. “Mostly they are fine being in a home environment by themselves. But parrots are not domesticated. They are wild animals, and we shrink their world to the size of a single room in your house, or even to the size of a small cage. We have to remember that. That’s why we need ways of entertaining them when you aren’t there.”
The show then cut back to the Lab, where Wart was expertly working with the Serial Tr-Hacking equipment. I then described what we called the InterPet Explorer. Ben Resner had suggested the idea at an early brainstorming session on finding ways for parrots to entertain themselves. “Why not find a way for parrots to be able to surf to sites they find interesting?” He meant it more as a joke than a serious idea, but the Media Lab people loved it, and we instantly got funds for developing it. When Ben finally had something up and running, it wasn’t really a way for Alex and friends to surf the Web, though journalists loved that idea. Headline writers had a lot of fun: “Give That Bird a Mouse” was one example; “Polly Wanna Web-Surf” was another; “On the Internet, No One Knows If You’re a Parrot” was yet another.
The idea ultimately came down to Alex, or whomever, being able to select between four modes using a joystick: pictures, music, games, and video. Within each mode were four choices. In music there were clips of a classical piece, a rock song, jazz, and country music. Wart came to be quite proficient at surfing the availabilities, as did Alex. But Griffin showed little interest. Our hope was that a bird alone at home would be able to entertain itself for hours with its favorite activities. One practical problem was to generate a large enough number of options. How many times would you want to listen to the same fifteen-second clip from Vivaldi’s Four Seasons before you’d want something else?
At the time we shot the Frontiers show, the prototype InterPet Explorer had just been put together. Alex had had little experience with it. As Alda said, “During filming, Alex remained resolutely uninterested in the browser.” That was no surprise to me. Alex was much more interested in Alda and me, and the rest of what was going on around him, than in this limited little machine. So we decided to film him in front of the browser when there was no one else in the room. He did pay attention to it, though he was not very interested in the pictures. Instead, he loved the music. The Frontiers show ended with Alex perched in front of the screen, repeatedly selecting the music option, and enthusiastically whistling along with whatever was being played, head bobbing, obviously having some private fun.
Everything about the Media Lab was such a great improvement over my situation at Tucson, except it was cramped. I had an office of my own, which I shared with Wart. He was regarded as something of a Lab pet. He was not alone in that role: there were several dogs wandering around.
Our work space, just off the third-floor Pond, measured about ten by fifteen feet—not a big room. Ben had a desk in there, as did Spencer Lynn. Spencer had been a grad student with me at Tucson, where he had been Alex’s absolute favorite person. I had a special relationship with Alex, obviously, but in general Alex preferred guys, especially tallish guys with longish hair, like Spencer. Alex often would pad around the Tucson lab, looking for Spencer. When Spencer picked him up, Alex would run up his arm, perch on his shoulder, and perform the Grey’s mating dance. Spencer was the only person Alex called by name. He used to say, “Come here, Ser.”
In 1999, however, Spencer had committed a cardinal crime, in Alex’s estimation, when he went to Africa for three months to study the behavioral ecology of Greys. He had abandoned Alex. And Alex never forgave him: Spencer was no longer his out-and-out favorite. And now they shared a small room at the Lab, together with Ben, Griffin and Wart, a temporary programmer, several computers, all kinds of other electronic equipment, soldering guns, two bird cages, and a troop of undergraduates whose job it was to train the birds. I don’t know how Ben and Spencer ever got anything done. They would sit at their desks, trying to read, work on their computers, or construct something for one of the Pet Projects. Within a few feet of them were the undergraduates saying things like, “Alex, what color three? What color?” and “Griffin, what matter?” When the birds weren’t answering questions, they might be calling loudly.
It made for quite a bizarre scene in this temple of high technology. In the end, Ben and Spencer resorted to some low technology to save their work sanity: they bought industrial-strength hearing protectors, the kind used by people who guide airplanes to their docking positions at airport terminals. Ben said it worked. They could do their work, the cacophony around them reduced to a low hum.
Around five o’clock each day, Alex said, “Wanna go back. Wanna go back,” signaling that it was time for him and Griffin to be taken to the animal care facility, where they spent the night. Wart, meanwhile, came to my office, where he spent the night. “Want to go back” was to be heard elsewhere, too, from Ben’s lips. He told me that partly as a game he tried to see how far he could get using Alex’s vocabulary. Not far, it turned out, but Ben did find himself sometimes saying “Want to go back” at the end of a party with his wife, for instance. His friends thought he was crazy, as did a waiter on one particular night in a restaurant.
The waiter came up to our table and said, “Would you like to hear about our specials?”
We said, “Sure.”
He sai
d, “We have Chilean sea bass in a pesto sauce, served with squash and green beans.”
My wife and I looked at each other and did the “green”…“bean” duet that Alex and Griffin often did. The waiter gave us a Are these guys nuts or what? kind of look.
Mr. A. did have a way of getting into people’s heads.
The fall 2000 sponsor event was scheduled for a few weeks after the Frontiers filming. I hadn’t planned to have Alex demonstrate what we were doing with phonemes, the individual sounds that make up a complete word, but sponsors had asked to see that task. We had started this project in Tucson and continued it at the Media Lab. We were training Alex to sound out phonemes, but not because we wanted him to read as humans do. Instead, we wanted to see if he understood that his labels are made up of sounds that can be combined in different ways to make new labels. We knew that he sometimes babbled when alone, producing such strings as “green, cheen, bean, keen,” and so on. This suggested that he did indeed understand that labels are made of subunits that can be used in different ways. But, as always, we needed more scientific proof.
We used plastic refrigerator letters, each a different color. We taught him the sounds of the different letters or letter combinations. We would ask him, for example, “What color is ‘ch’?” and “What sound is purple?” He had become quite proficient.
Alex & Me - How a Scientist and a Parrot Discovered a Hidden World of Animal Intelligence--And Formed a Deep Bond in the Process Page 12
