Chapter 5

Laura did her best work early in the morning. She could, and often did, work for the whole day, but she never had that crisp sense of inspiration, of being at her peak, any time after noon. She'd now shared a room with several different people, and she understood not everyone worked quite like that, but it still surprised her at a certain level. She was always bemused when Amanda would toss in bed, murmur some sleepy noises and go back to sleep as she left to work, and at the start it had given her a bad opinion of her. It seemed like a lazy thing to do. Of course she had since seen Amanda working as hard as anyone, but for Laura wasting the morning seemed like an offence against nature. At least Amanda was considerate, unlike one of her previous roommates, who insisted on doing much of her work at night, while listening to infotropic background music.

She headed to the lab, and wrote a quick comm to her supervisor: 'Since we couldn't talk about my work yesterday, please slot a meeting ASAP. It would be good if you could skim my last progress report.' It was fine to reach a deal with her supervisor, but it wouldn't do her any good if she couldn't actually help, and the sooner she could find out the better. Laura didn't think of herself as vindictive, but she detested rule breakers, and if it weren't for the problems she had been having with admin she would have simply stormed out of the room and told everyone about the pervert.

At the lab, she fiddled with some of her models, as she waited for one of her subjects. His name was Adrian, and he was a rather old guy, for a student, in his late 20s. Laura didn't particularly care for Adrian personally, as she didn't for any of her subjects (they were just data sources, as far as she was concerned) but Adrian had a unique feature: her models actually worked for him. She had been presenting him with poems and asking him to rate them, and her algorithms were capable of predicting his ratings from the brain states she could measure while he read. Why such a model would work with one subject and fail, to a greater or lesser extent, with most of the others, was something she needed to get a grip on.

Eventually, Adrian arrived, as usual a few minutes too late, which was yet another reason why Laura couldn't warm to him, and she set her equipment up.

"Ok", she said. "We'll try a different protocol today. Instead of reading poems, I will play you some music. After you listen to the piece, I will ask you to rate it from 1 to 5."

"Fine, but I warn you I don't know much about music", he said.

"That's not a problem. I only want your subjective evaluation of the music."

"So is this a study about taste formation or something?"

That was another reason why Laura didn't like Adrian. In fact, she disliked him more and more the more she needed him to verify her models, and the more he kept asking questions he knew or should have known she shouldn't answer. Depending on such an unpredictable person annoyed her, and to Laura everything about Adrian suggested unreliability: he studied some recondite pointless specialism in the humanities no rational person could care about, as it was hopelessly fuzzy; he dressed in a way that seemed designed to make the point he didn't concern himself about his appearance, but it was obviously an affectation; he was invariably late. It was almost as if fate had chosen in Adrian the perfect way to offend Laura's sensibilities, and in a way she could not avoid, at least until she had equal predictive power for other subjects.

She sighed. "I cannot confirm nor deny whether this is a study about taste formation."

"Oh, sorry, never mind me. Go on then."

Laura placed a band around Adrian's head, annoyed at his long hair. The band was full of small electromagnets and sensors, and had a short and rather rigid finger-wide cable coming out the back, connecting it to the deconvolution system, known as a devolver by those who used it, and who tended to have an ambivalent attitude towards it. It was a metal cube, with a small screen on one side, where Laura had to conduct her readings and adjustments, and four pipes, one on each corner, which took care of the coolant.

She wasn't that interested in the arcana of hardware but she had heard the system consisted of 256 thin cards that were too dense and complex to be fabbed, surrounded by copious amounts of piping to make sure they would never overheat. If the rumours she heard were correct, a devolver cost as much as the lifetime wages of ten professors, and there were constant arguments on whether it had been money well spent.

"Ok, at this stage of the protocol I don't need to supervise the system constantly. I'll be doing some work while you do your part. If you have any doubts or there's any sort of problem, call me."

At least she didn't have to stand at his back, tweaking the interface as she asked aimless questions. Time passed quickly. She got both comms by Amanda, and managed not to betray any reaction. She found the idea of someone masturbating in her room disgusting, and yet somewhat intriguing. Not that she'd ever tell anyone that. Not much later, her supervisor commed her: 'Today I am free in the afternoon. I can see you have lab work in the morning, but you could come by then if you wish. I'm familiar with your reports, believe it or not.'

Soon Adrian finished his task, and Laura made sure to thank him for his cooperation. As she got ready to deal with her next subject, she took a look at the correlation between her rating predictions and the actual ones, expecting to see the 1.00 she had got used to from the previous protocol. It was 0.54, little better than random. She couldn't believe it.

"That is just wonderful", she muttered. "Just. Fucking. Great."

Amanda had never written a proposal to the fabbing board, and she wanted to make sure it would be just right. She was certain Professor Robertson would let her know if there was anything wrong, since he was going to co-sign it after all, but she had never been one for turning in substandard work, so she tried to do her best, a bit hampered by her lack of experience. She wrote of her work's potential to improve clinical diagnoses, how it could be used to monitor the effects of treatment in realtime, measure the changes in blood composition, etc. She had the feeling there was something she was missing, but she couldn't put a fingner on it, so she just kept on, and moved onto the feasibility part of the proposal.

Fabbing, as everyone kept saying, had changed the world. As with so many other technologies, it hadn't quite lived up to the expectations of its proponents, who thought all objects would be custom-built and free, and that everyone would simply fab and recycle them with a unit at their home. That said, it had definitely gone further than its critics suggested. The reason why physical objects hadn't become software-like, copiable and updatable, was due to fundamental technical restrictions of fabbers: true, you could have a fabber for a week's wages, but it would be slow, and the objects wouldn't be that polished. Proper fabbers, like the ones the university had, could print a cubic metre of a single solid material in an hour and they could make objects with any number of different material layers, down to the precision of 5 micrometres. A unit like that cost a lot of money, however, and although progress hadn't stopped, it seemed that it would always do.

There were always more fabbing requests than there was fabbing time. Amanda needed at least 10 units to begin trials, and although her new design was a lot simpler, fabbing the microcircuitry, sensors, etc, would take up quite a while. She wasn't making a simple solid, but a very complex artifact, with a huge surface and many different layers of different materials. The simulator estimated building one of her rings would take up 5 hours, though it wouldn't use the total capacity of the fabber, so other things could be printed concurrently. It was a design that required little space, but a lot of extruders and depositers.

As she kept trying to justify her use of the machine, she got a comm from Lenka. She had been working on the proposal for a while, and she felt she needed a rest, so she read it.

Hi, Amanda. Sorry about yesterday. I think I did truly shock you, which is what I wanted, I suppose, but I didn't mean to make you feel awkward. Astrid says I scared you away for good, but I hope it wasn't that bad. Really, I didn't have a clue you were going to tell me about your own situation, or what it even was. It was perhaps an unfortunate coincidence.

Astrid and I would like to make it up to you, if possible. We'd like you to visit us at my place, and I promise I will be on my best behaviour. We've been friends for a long time now and I want to keep it that way. Let me know when you can make the time, please, and if you want to make it a meal.

Sorry,
Lenka.

After she finished reading, Amanda was a bit surprised. Her reaction must have been a lot more obvious than she'd thought. It was true she had felt shocked, and just as true she had both found it alluring and a little disgusting. She didn't want to see her own future in Astrid, crawling on the floor of a pub just because someone, on a whim, told her to. Still, they had got the wrong impression from it all: she didn't blame them for it, she just found the situation difficult to deal with.

She did want to meet them again, though. She thought she could maybe prevail on Astrid to tell her more about what it was like to be a sat. She felt so torn: both repelled and inescapably attracted to the idea. Of course it was all silly nonsense: Laura was as asexual as they came, and the least likely person to want a lesbian liaison. Even if she was obviously a take charge kind of girl.

She thought about it for a bit, and in the end wrote them both a comm.

Hey, Astrid and Lenka. Got your note. Don't worry, you did give me a shock, but I wasn't that put off by it. It was just surprise and, well, uncertainty. I had seen things like that before around you, so I suppose I should expect it by now. Still, I admit it had a big impact.

Since you're offering a meal, though, and me being a growing girl and all that (especially in girth!:-)) what about tomorrow lunch? Is it a date? ;-)

Laura's morning kept getting worse and worse. Her work with other subjects required her to adapt her questions to their responses, and to the readings she was getting, so she had to stand up behind them looking at the small screen on the devolver. This, needless to say, wasn't a natural way of talking to anyone, it distracted the hell out of her subjects, and made her really tired and grumpy. By the time her session was over, she was feeling annoyed with the world, and was growing more and more convinced that she had run against a wall. The negative results she got from Adrian really shook her, and she wondered if she may have chosen too ambitious a goal.

After she sorted out her equipment and backed up her notes, she went to see her supervisor. If she didn't have any ideas, she was going to have to think things through very carefully indeed. It took her just a few minutes to walk from the lab to the office. She got there in a bad mood, and reminded herself her current difficulties weren't her supervisor's fault.

She knocked, and Professor Hyde called her in.

"Hi", she said. "Take a seat. What would you like to discuss?"

"The fact my models aren't working. I seem to be able to make them work if I get some calibration, but they rarely get accuracy over 80% and there just has to be a way to avoid having to go through this stuff."

Laura went on to summarise the different models she had tried, and the difficulties she kept getting, like the variability between subjects.

"You know people are wired in very different ways", said the professor. "I am not surprised that you can't model everyone that easily."

"Yes, but it works sometimes. What is going on? Today I had a really weird result. I'd been having perfect prediction with Subject 8. I changed the protocol in a completely irrelevant way and my prediction is little better than random."

"Hmm, that is actually good."

"How so?"

"Think it through. You had an accurate model with the subject, then you did something and the model no longer was accurate. Why? What change did you make?"

"On the first protocol the subject was rating poems, and on the second rating songs. It seems like the same task, and what my model suggests was being measured was aesthetic appreciation."

"You need to allow for more dimensions then. Perhaps emotion-space doesn't let itself described in scalar terms. Another emotion may have been confounding your output."

"That sounds ... hmmm, plausible. If we assume emotions aren't independent, but instead add up by some kind of vector operation..."

"This thing with calibration ... you won't need it in the end."

"How can you tell?"

"Because you didn't need it with Subject 8. There's something clearly and fundamentally correct about what you're doing. It's just getting hidden under other variables. I think you need more input to resolve it."

"More input? The devolver is at its maximum resolution."

"I don't mean that", said Professor Hyde. "The devolver is sufficient if and when your model is correct, but to find out what is going on, causally, you're going to need different inputs. I don't know what they are yet ... but I can feel that you will find it very difficult to see the underlying process without a different data source."

"Any suggestions on sources?" Laura sounded impressed in spite of herself. She found it hard to comprehend how Professor Hyde could be an excellent specialist and an idiot who messed with her students at the same time.

"No, I just don't know. You need to think outside the devolver, though, use some data which is causally dependent on similar phenomena but separate from brain waves."

"I'll look around, see if there's anything useful in the literature. I'm starting to feel like I won't solve this one."

"You'll do it. You're stubborn enough. I wouldn't use the literature except negatively, though. I have the feeling you need something new."

"Negatively?"

"Right", said the professor. "As a list of things not to try. I don't think you're going to find what you need has been done before, somehow."

"So I have to become familiarised with a whole different discipline to get this done?"

"That's how it looks from here, but if you disagree, well, I could be wrong."

Laura had been thinking about it, and what the professor said seemed to resonate with her own intuition. There was something at the back of her head, something she couldn't quite express. She thought she had the answer if she only could remember it, somehow. At any rate, it seemed her supervisor was certainly worth keeping around.

"Well, I get the feeling you're right. Anyway, it seems like the rational thing to do at this stage. Read the literature, try things out, attempt to find things that haven't been done before. I guess I should thank you, you were helpful."

"I guess you should. Let me know if anything comes to mind. If I turn up anything useful I'll pass it on too."

After having a small lunch, Laura went back to the lab to perform more experiments. What had happened with Adrian had confused her. Aesthetic appreciation should be the same whether one is confronted with poems, songs, or members of the opposite sex, or for that matter of one's own, given people like Laura and her friends. What Professor Hyde had suggested seemed to make some sense: emotions are complex, and they may interact with each other in strange ways. Yet somehow she didn't think it was quite that. Maybe that was a factor, but it couldn't be the fundamental problem.

After a while, the routines of her experiments drove all other ideas from her mind. She didn't have the energy and focus to think about her problem, as she kept asking questions and measuring responses, feeling more and more tired of standing up behind her subjects. Her feet were uncomfortable again, and although she began to move a little, in small steps, always careful not to lose sight of the screen on the devolver, it didn't seem to help a great deal. Perhaps she could convince her roommate to give her another foot rub when she got back. Somehow, Amanda seemed a little peculiar about it, not in a way she could put a finger on ... just ... Whatever, she didn't have the time to think about such things then.

By the time she finished her work, there wasn't much light left. She was almost, though not quite, as tired as she had been that first day she had been allowed the use of the devolver, and all she wanted was to sit down, have dinner, and go to sleep. All ideas about her problem, or about Amanda's behaviour, were beyond her grasp, tired as she was. She made sure to put everything away, backed up her data to her comm, closed the lab, and trudged home.

The whole day had gone, and Amanda hadn't done any actual work. Well, work in the sense of useful, tangible stuff. Of course writing the proposal was just as important as anything else, but it felt unproductive. She was still fiddling with the proposal, trying to make it better, when she got a comm from Laura: 'Really tired today. Heading home. Could you get started on dinner, please?'

Laura could be so rude sometimes. Or was it rude? It was true they had reached an agreement that Amanda would do the cooking, but asking her to have dinner ready ... It was like something out of the industrial era, when half the workforce did all the housework for the other half. But perhaps Amanda was just being oversensitive: Laura had had a difficult day, standing awkwardly by her equipment, and just wanted to have a warm meal and rest as soon as she got home. Not too unreasonable, when put that way, was it? Plus Laura seemed to have a weird notion of politeness: it wasn't that she was impolite, as clearly she tried to be considerate; it was more the case of having been exposed to very different norms. As if she were foreign, or a time traveller. Perhaps that's how Amanda could think of her: as someone from a different culture.

After a bit of dithering, while Amanda considered comming back and telling Laura she wasn't her maid, acquiescence won out. What was the point in having an argument with Laura, when in all likelihood Laura hadn't even realised what her comm sounded like? It was time to get dinner ready, anyway. Amanda was getting slightly hungry herself, so there wasn't much to argue about. Just the forms, and Laura had even written please.

At the kitchen, Amanda took a look at her supplies, and decided to do something a little more elaborate than a soup. She wasn't going to make anything too complicated, just some fried chicken with garlic. A bit more involved than warming up lentils, but reasonably quick. As she worked, something in her responded to her situation: she was cooking for someone else, taking care of Laura, in a sense. She was serving her. Of course she was cooking for herself too, but Amanda wasn't used to doing it for another person, and she somehow liked it. It felt ... useful. It made her feel needed, wanted, and the fact she was obeying Laura's comm warmed her somehow.