Anyone want to go through the September LSAT questions?

Comments

• __Juan__ Alum Member

  October 2016 184 karma

  It says people with high homocysteine levels are twice as likely to be diagnosed for Alzheimer's. It then claims that indirectly reducing homocysteine levels with vitamin B can reduce the likelihood of Alzheimer's.
  
  But if E is true, Alzheimers tends to be the cause of an increase in homocysteine levels. So, reducing homocysteine levels won't reduce the risk of Alzheimer's. It's possible that those who had that high homocysteine diagnosis already had alzheimer's and that's the reason their levels were high.
  
  D talks about relatives and that was never mentioned in the stimulus. It has nothing to do with the relationship between homocysteine and alzheimer's.
• nomi.jeanne Free Trial Member

  October 2016 6 karma

  Yes yes please
  
  I need help clarifying LR 1/q 24.
  
  I understand why B is right - if the info in the computer program codes for the structures of the proteins but not their interactions, then a computerized model of the human genome couldn't perform the operations of the human brain. Therefore, scientists will require more than the info in the human genome to create AI.
  
  I can't succinctly explain why A and C are wrong. I chose A on test day, and C on blind review. I probably skipped over B because it's worded confusingly >.<
  
  I'll do my best below, someone please tell me if I'm understanding this right. Been thinking about it for half an hour.
  
  A (negated) 'The functions of the human brain are governed by processes that can be simulated by a computer.'
  The biologist's argument could
  still be valid. Even if the functions* of the human brain could be simulated by a computer, it will require more than a computerized version of the human genome to mimic the human brain if those functions aren't coded for in the genome (essentially what B is saying).
  
  Here's what confuses me: are we supposed to assume that modeling the operation of the human brain is necessary for creating AI? I know that sounds silly but this is the LSAT, take no common sense for granted.
  
  *I'm assuming that functions = operations.
  
  C (negated) 'there are other ways to create an artificial intelligence besides modeling on the operation of the human brain.'
  
  Here's where I got f'd up. This answer leaves open the possibility that there are other ways to create AI using the human genome and modeling it on something besides a human brain. I know that sounds crazy when it's written out in plain English, but like I said earlier, are we just supposed to assume that we're modeling an AI on the operation of a brain? And isn't that sort of a necessary assumption? I'm a little rusty after nearly a month off of studying, but bear* with me here. If the scientists model AI on something else, couldn't they still use the human genome and nothing else? Why do you need a brain model at all?
  
  (And while we're on the subject of bears, I will admit that I had to read question 23 five times just now before I realized it compares dead bear bones to living bear blood. Honestly I feel like I deserved the 8 point test day drop. The answer was immediately obvious after that. Sloppy.)
• mk940808 Alum Member

  October 2016 209 karma

  @nomi.jeanne
  
  So the way I see that question, the difference between what the computer scientists think and the author thinks is that the operation of the brain is governed by proteins whose structures are ENCODED IN the genome, but not of the genome itself. So the question assumes something regarding that small difference in wording and we have to look for an answer that, if negated, will mess that up.
  
  A is irrelevant, the author isn't saying that the computer will never be able to simulate it, just that it can't be simulated merely by encoding the genome. So the negation doesn't affect the argument. Also, this is a little off topic, but the way the author set up the argument sounds like s/he will end with "so the only way to do AI is to simulate the proteins encoded in the genome" or whatever. So just based on the way that it seems to be going I would be a little suspicious of anything that said NO AI IS POSSIBLE straight up.
  
  C does nothing to address the difference mentioned and is somewhat irrelevant -- the author didn't mention "the only way" so you can easily toss it out. It's an extreme answer, and negated it doesn't affect the argument - "there is not only one way to do AI"
  
  B is correct because it addresses that difference and the entire argument would be negated without it. If it is true, then the scientists are right and the argument is wrong.
  
  
  And yeah the bears!!!! Argh!!! Only realized that after getting it correct the second time around -.-
• 1 Like

  mk940808 Alum Member

  October 2016 209 karma

  @nomi.jeanne I had a 6 point test drop -.- so I feel you, I'm alternating between rage and despair.
• nomi.jeanne Free Trial Member

  October 2016 6 karma

  @mk940808 - I think it's possible that you're interpreting this correctly but I'm still confused on the difference between 'encapsulates the information' and 'encoded in.' To me, 'encapsulates the information contained in' is the same thing as saying 'encoded in.' They're both referring to the information contained in the human genome.
  
  The way I interpreted it, the biologist is saying that the coded information isn't enough to make AI. Something has to trigger the interactions between proteins in order for the brain to operate - and that trigger isn't inherently part of the human genome. This is a really dated example, but I think it would be equivalent to saying 'all you need to watch a movie is this tape.' All the information is encoded on the tape, but without the vcr, you can't watch it.