CogSci1 Top 10 List Chapter Summary
xex Home

The following Top Ten List of Big Ideas summarizes Chapter 5, "The First AI Debate", of Stan Franklin's [[Artificial Minds]]. The the words are generally taken straight from the text.

  1. Computers Thinking Like Humans A very important question when it comes to AI is if we can expect computers to think like humans.John Haugeland states: "The fundamental goal of this research is not merely to mimc intelligence or produce some clever fake. Not at all. "AI" wants only the genuine article; machines with minds, in the full and literal sense."

  2. Scoffers John Haugeland also states that "scoffers find the whole idea quite preposterous - not just false but ridiculous - like imagining that your car (really) hates you or insisting that a murderous bullet should go to jail."

  3. Boosters John Haugeland has a bit to say about boosters as well. Haugeland says "boosters... are equally certain it's only a matter of time; computers with minds, they say, are as inevitable as interplanetary travel and two-way pocket TV."

  4. The Turing Test The question is how can we tell if a computer has actually achieved AI, the answer is through the Turing Test. With a Turing Test a barries separates an interrogator from a human and a machine. The interrogator would then ask questions via a terminal to both the human and the machine. If the interrogator cannot reliably tell which responder is human and which is machine the the machine has achieved AI.

  5. The Pro Position Simon makes two statements to support AI. These statements are that "intelligence is the work of symbol systems" and "a physical symbol system has the necessary and sufficeint means for general intelligent actions. The computer is a physical symbol system. The most important is the human mind and brain."

  6. The Dreyfus Attack Dreyfus states that "the physical sybol system approach seems to be failing because it is simply false to assume that there must be a theory of every domain." Dreyfus claims that not every domain is rule describable.

  7. The Chinese Room The Chinese Room focuses on the problem of understanding. John Searle created the Chinese Room problem in 1980 and experimented on himself. He placed himself in a closed toom and then a sotry and a question about it, which are written in Chinese, and slipped under the door. Searle has no basic understanding of Chinese, the only thing he has to work with is a comprehensive sript that provides an algorithmic way of answering the question as a native speaker would. Searle is able to provide the proper answer for the question he was given, which would imply that he understood the script and the question. Searle uses his Chinese Room experiment to argue Schank's notion of a script.

  8. Scriptin' like Schank Scripts are used for representing knowledge of common sequences of events. Schank creates a script and then is able to ask a program a series of questions based on the script. If the program can successfully answer the questions correctly then it is supposed to mean it understands the script. Searle cleary shows that this is false with his Chinese Room experiment.

  9. The Penrose Attack Penrose claims that computers operate algorithmically, as do brains, but unconsciously. He states that consciousness evolved to make judgements that cannot be made algorithmically. Humans make decisions that are no made algorithmically and therefore cannot be made by computers. The Penrose attack is a two pronged attack: the nonalgorithmic nature of mathematical thogh and quantum mechanical effects in the brain.

  10. The Horgan-Tienson Attack Horgan and Tienson both have a theory that says machine intelligence cannot be produced via symbolic AI alone with rule-based systems. Their attack is a three pronged attack focusing on multiple soft contraints, cognitive folding, and the frame problem.