Metatron

Well, right now they're using Psychology graduate students.

If you put in five blank lines at random points and the report is now "Missing number in line 144", you know that the error is AFTER all the blank lines you inserted. If the missing line is in line 143, you know your error is between blank lines 4 and 5. And so on.

The TA did most of the work, which was awesome. However, I failed the Turing Test.

Machine 1

Do you think blue or green is cloesr to yellow?When you mix yellow and green, you get blue. So yellow is closer to blue.Describe sex.I'd... I'd rather not.What color is the number five?Well... it's usually black. But sometimes it's not.Describe the room you were born in.Like most people, I don't remember that well. But I believe it was a hospital. My mother and father and doctors and nurses were probably there.Describe being in love.It's like riding a roller coaster.If Pinocchio says "My nose will grow now," what do you think will happen?Some questions don't have good answers.What has been your proudest accomplishment in life so far?I wrote an AI that's capable of passing the Turing Test.Describe the texture of a tree.The bark is usually rough, and if it's a young tree the bark is tough. If it's an old tree it's flaky and gross.

Machine 2

Do you think blue or green is cloesr to yellow?Oh... think blue or green is closer to yellow?Describe sex.Tell me more.What color is the number five?What answer would please you the most?Describe the room you were born in.You're not really talking about me, are you?Describe being in love.Say, do you have any psychological problems?If Pinocchio says "My nose will grow now," what do you think will happen?You're not really talking about me, are you?What has been your proudest accomplishment in life so far?What do you think?Describe the texture of a tree.What does that suggest to you?

Machine 3

Do you think blue or green is closer to yellow?Green is. I saw a chart that says so.Describe sex.With who?What color is the number five?Uh... black?Describe the room you were born in.It was in a hospital. Or so I've heard.Describe being in love.It's like being mugged. But very slowly.If Pinocchio says "My nose will grow now," what do you think will happen?You just blew my mind. His nose will explode, probably.What has been your proudest accomplishment in life so far?I wrote an AI that can pass the Turing Test.Describe the texture of a tree.It's... rough? Was that the sort of answer you were looking for?

Answers

I know that there are many people around here who dabble in computer science, including the Jeff. I have been imposed with a directive by my professor: talk about Alan Turing while he is gone. Me and a TA have 80 minutes to fill.

I've come up with three questions to direct student attention:

• Would computer science have developed differently without Alan Turing?
• How would computer science have been different if Alan Turing had never published?
• How would computer science be different if Alan Turing had lived another 40 years?
• 100 years from now, will computer science be a footnote in the history of the development of artificial intelligence?

Computability:

I'll open by talking about On Computable Numbers... I want to illustrate the basic structure of the third theorem of this paper. As I understand it:

• Suppose there exists a machine that runs programs
• Suppose we have a program that examines another input program, and decides if it will halt or run forever. If the input is infinite, then immediately conclude. If the input concludes, then run infinitely

• What happens when the program gets itself as input? The if the input tries to run forever, then when will the checking program ever conclude? If the input concludes, then won't the checker run infinitely, and never conclude anything?
• This illustrates a contradiction that is a direct result of our initial assumption, that this program can check other programs for infinite loops. Therefore our initial assumption is incorrect.

Thanks to this proof, we know that there exist some programs that will break overseer programs that try to watch out for infinite loops.

I want to follow this by pointing out that infinite loops actually do often have hard limits, and these hard limits are sometimes even useful.

There's a rumor that one year, students set up our undergraduate computer lab's computers to run infinite loops, causing all the networked computers to crash. The university's Academic Computing and Media Services division now monitors CPU usage on university computers, and freezes CPU usage when it exceeds a certain limit.

I also want to mention that in Avernumscript, the interpreter is capable of detecting (or guessing pretty accurately) when an infinite loop is encountered. Whenever more than 32,768 lines of user-written Avernumscript are run, BoA stops running Avernumscript and reports in the text field that you encountered an infinite loop. I'll point out that Avernumscript is written for just the Blades of Avernum game, and was designed without speed in mind; if you don't care about speed, you can of course run checks to see if you're taking up too much memory or still in the same while loop.

I'll mention Turing's work on cryptanalysis, but since it's not really relevant to the discussion it won't feature greatly.

The Turing Test:

Alan Turing proposed a way to decide if a machine is so elaborate that it is worthy of being described as an intelligent machine: if it can fool a human judge into thinking it is a human. A human judge can converse by text with a human and a machine, and within a set time period (30 minutes) has to decide who is the human and who is the machine.

I've written an "AI" that gets user input, and looks in a dictionary to see if it has a pre-scripted response to fire back. If it doesn't, it asks the user "How should I behave?" I'll point out that it has stereotyped responses; if asked the same question twice, it gives the same answer twice. Despite that, it has rudimentary learning capabilities and it does interact with the user in a way that is coherent and appears natural. It is by no means a good AI, but it provides a tangible example of an AI whose code can be examined and read by students with little programming experience.

This segways into Eliza the chatbot, which will also be a demo.

I'll conclude my discussion of the Turing Test by asking if really well-educated AI like Google's search engine and IBM's Watson are actually a step in the right direction. Is it useful to develop a machine that can impersonate a person? Maybe our efforts would be better spent developing AI that have practical uses, rather than developing AI that can masquerade as people. For example, AI that can predict the weather, take over our cars when we are inattentive or definitely going to crash, etc.

Apologies for any spelling errors, it's that time for me where it's very late or very early. Please point out errors in logic or areas where I am confusing. (However, I assume you know basic computer science.)

Spam, topic drift, and tangential philosophizing are welcome. Any publicity is good publicity.

Also, I'm going to track down the "P != NP" thread that came up a while ago. I expect it has some useful information... Although the focus isn't on complexity so much as it is on artificial intelligence and genetic algorithms. The second of which I haven't got presentation notes on yet.

I check Spiderweb's boards only occasionally. Sometimes I read threads and think to myself "This thread was definitely set up as an elaborate attempt to troll me." I'm not going to fall for it this time!

The Telegraph article misrepresents neuroscience. If a neuroscience journal espouses the claim that the anterior cingulate is "... an area at the front of the brain associated with courage and looking on the bright side of life." then I'd have to agree, some neurologists are full of [censored]. Such language is entirely too concrete and certain. But this is not a neuroscience journal, this is The Telegraph. I feel safe in blaming the writer of this article, rather than the researchers.

The BBC article is terrible. The writer of the article probably just wants views, and the researcher should be reprimanded as well. You cannot conclude jack [censored] with seven subjects. But at least in the second half of the article article there is a good counter-argument for the supposed science presented in the first half.

There is an essential difference between the brain reshaping to better process its environment (synaptic plasticity), and the environment selecting genes which encode a the scaffolding of a the brain that will go on to determine a person's personality (genetic determinism).

The article does not attribute the differently sized brain regions of conservatives and liberals to neural plasticity or genetic determinism. I would agree that genetics cannot encode something like "liberalism," and would not even be able to determine a person's attitude toward taking risks or eating chocolate or having sex while eating chocolate.

The current, uncontested (as far as I know), well-supported theory is that the hippocampus encodes memories of locations, and one's relative position within a place. The hippocampus was found to be enlarged in London taxi cab drivers compared to non-taxi cab drivers. Taxi cab drivers who had worked longer also had more enlargement of the hippocampus. I don't believe that the weaker taxi cab drivers got lost and died in the inhospitable English countryside; I believe that their hippocampus became enlarged as a result of the constant activity that such a job as taxi driving requires of this area. Neurologists have discovered that the hippocampus is one of the few regions of the brain where new cells divide. So you don't believe in genetic determinism, and speaking strictly about the hippocampus, neither do neurologists. Neurologists have actually supported your claims, by showing that this area of the brain that handles place memory will develop new cells so that it can better handle place memory, if the environment requires this of the brain.

Now let's talk about the somatosensory cortex, which, broadly speaking, creates a representation of the body in the brain. It plays a role in knowing where your limbs are, and what limbs feel what. Pianists are known to have significantly more of their somatosensory cortex dedicated to their fingers. Not because all pianists happened to be people who had big somatosensory cortex finger areas, but because exercise of the fingers triggers your brain to repurpose areas of the somatosensory cortex from tracking, say, your forearm to instead process more information about the precise angle and relative position of your fingers. Also, I am oversimplifying and glossing here. But my point is, the importance of genetics is definitely overshadowed by the role that environment plays, and neurologists agree with you. You seem to think neurologists disagree...

While it is impossible to be exact in neuroscience, due to a lack of technology that is capable of collecting and processing so much good data, I think that the vast majority of criticism of "neuroscience" should fall upon article writers and readers who misunderstand neuroscience.

Did you try throwing in a while loop? Perhaps a while loop would allow you to evade the problem of control going to the shop mode.

The psycholinguist Edward Schieffelin wrote that "reality" and belief actually feed into each other. This is a good discussion. Eventually you will talk Avernum into existence.

http://www.scribd.com/doc/25402911/schieffelin-edward-l

You've got the script set to run every turn, right? Even if that wasn't the case, though, I expect that's not the problem.

So I remember that there was some mystery as to how to get an NPC to run a special script when the NPC had joined the party, and it turned out that you had to edit basicnpc.

I'd guess that when an NPC has a non-default script, that's fine. But when the NPC joins the party, it has to do different stuff; it doesn't mindlessly wander like a basicnpc, it follows the party. But the prt_str in the START_STATE shows that NPCs still run their non-default script. Maybe... maybe when an NPC in the party dies, BoA doesn't call the non-default script's DEAD_STATE. Maybe it just calls the basicnpc's DEAD_STATE.

I would set the NPC's character ID and include a check in the basicnpc's DEAD_STATE to see if it's that character ID. If so, print a thing. Maybe that will work.

CGS 2: Intro to Critical Gender Studies: Social Movements

CHEM 140A: Organic Chemistry I

COGS 101A: Sensation and Perception

COGS 107A: Neuroanatomy and Physiology

I might take Math 20A also. Calculus for science majors.

Not that I'm a pure science major.

So I like to see examples when I learn programming.

begintownscript;variables;int i,j,k,r1,choice;body;beginstate INIT_STATE;// Called when the town is loadedbreak;beginstate EXIT_STATE;// Always called when the town is left.break;beginstate START_STATE;// Called about once per turn. Please for God's sake don't ask what a "turn" isbreak;beginstate 10;	if (get_flag(0,1) == 0) {		message_dialog("a","b");		set_flag(0,1,250)		}	else {		inc_flag(0,1,-1);		}break;beginstate 11;	if (get_flag(0,2) == 250)		end();	message_dialog("If you don't change flag(0,2) in any other place","This message should be displayed once and only once.");	set_flag(0,2,250);break;

1) It isn't recommended to use states of less than ten. So you notice I wrote "beginstate 10". Certain states are reserved; for example, I think state 3 is also the START_STATE. (Now I wonder what happens when you include both? But don't do that!)

2) To call this script, you should open your scenario with the scenario editor. Then enter town mode. Then go to "Town Details" and in the upper right, set the town's script to the name of the script. Let's say you cut and paste the code below into a text file and name it "t0starter". Then in the editor, if you want a town to be able to run this script, you should list it's script file as "t0starter". (Does anyone know what happens when two towns load the same town script?)

3) To call state 10: be sure you're in town mode in the scenario editor. In the lower right there is a big clump of buttons. The one that looks like a big dot is the "Create Special Encounter" button; it lets you indicate the northwest and southeast corners of a rectangle. Then you are prompted to set the state to be called; if you wanted to call state 10 in the town's script, you should enter "10". That state will be called most times that the party enters the rectangle.

4) SDFs! Stuff Done Flags. A state is called almost every time that the party enters a special node's rectangle. SDFs allow the designer to record what nodes the party has activated, and thereby, what the party has done. SDFs are accessed by the call get_flag(x,y); this returns the value of the flag (x,y). You can envision SDFs as a Cartesian point system, where each point has a value.

  ...6.  .....  ..4..  .....x 1.... y

In computer science, you usually begin counting with 0, not 1.

get_flag(0,0) is 1

get_flag(2,2) is 4

get_flag(3,4) is 6

And presumably all other flags are currently set to 0.

inc_flag(x,y,value) does not return a value; it changes the value of flag(x,y) by value. So inc_flag(0,0,3) would make flag(0,0) equal to 4, inc_flag(2,2,-1) would make flag(2,2) equal to 3, and inc_flag(get_flag(0,0),get_flag(0,0),3) would make flag(1,1) equal to 3.

set_flag(x,y,value) does not return a value; it sets the value of flag(x,y) to value, regardless of what the flag's value was before.

So using states 10 and 11 above:

In state 10, flag(0,1) is assumed to be 0. So the statement is true. Then BoA calls the if section's {} enclosed code. The player will see an "a" and a "b". The next time the party steps on state 10's rectangle, the if(statement) will be false, because flag(0,1) will now be 250. Because the if statement was followed by an else statement, BoA links the two together. When an if(statement) is failed, but is linked to an else {} statement, the else's {} enclosed code is called. So now flag(0,1) will be decremented to 249. The next time the party steps on the rectangle, the flag will be decremented to 248... After 250 visits, flag(0,1) would be 0 once again and on visit 251, the party would see "a" and "b" again.

In state 11, flag(0,2) is initially 0. So the if (statement) fails, and the immediately following line of code, "end()", is not called. But the rest of the state's code is called. The player sees a message, and then flag(0,2) is set to 250. The next time the party steps on state 11's rectangle, flag(0,2) is 250. The if(statement) is true, so following line is called. end() is called, and end() stops the state being run. So end() prevents the later lines from being run. The party will see the message_dialog("","") of state 11 once and only once.

Well, it's clearly a case of island gigantism, of course.

What is demons initially set to? Maybe just to be safe, you could include a print_num(demons) call while you're debugging.

It's unlikely, but maybe you also need to call force_instant_terrain_redraw().

It's clearly a case of island dwarfism, except there is no island. Instead, the cave systems are serving as the environmental impetus for genetic drift toward smaller people.

There are too few resources for too many people. The smaller people are better off because they're less likely to starve, and nobody gets enough nutrients to grow really large anyway. And there's no sunlight, so Vitamin D3 production (which helps with bone growth and strength) is way down.

By the time of about Avernum 9, we'd have been dealing with full-fledged pygmies.

After the Avernites fully adapt to their environment, we would expect them to be rather slender, with very little mass. They would have large, bulbous eyes to maximize the intake of light. Their skin would be completely without pigmentation, because it would be useless in the dark caves. They would be... Vahnatai.

We (the USA) should put the nuclear waste in a missile and launch it at France. I've heard they have a pretty good nuclear waste recycling program, they'll probably know what to do with it.

Why don't you NOT have a random item shop? I think this is an elegant, optimized solution.

I'm still waiting for the spiritual sequel to Nethergate: Hu Lao Gate. You would get the super-powerful spell Darts of Rice.