Upgrade to Pro — share decks privately, control downloads, hide ads and more …

class1-reduced.pdf

David Evans
September 03, 2018

 class1-reduced.pdf

Pavilion Seminar, “How will Artificial Intelligence change Humanity?"
University of Virginia, Fall 2018
https://aipavilion.github.io/week1/

David Evans

September 03, 2018
Tweet

More Decks by David Evans

Other Decks in Science

Transcript

  1. Plan for Today Brief Introduction to Seminar Introductions Humans Need

    Not Apply Snack break First major readings: Sapiens 1
  2. 5 Cognitive Task Human Machine (2018) Adding 4-digit numbers Adding

    5-digit numbers ... Adding 8923-digit numbers Spelling Sorting alphabetically Sorting numerically Factoring big numbers Playing chess Playing poker Playing go Face recognition
  3. 6 Cognitive Task Human Machine (2018) Adding 4-digit numbers ü

    Adding 5-digit numbers ü ... ü Adding 8923-digit numbers ü Spelling ü Sorting alphabetically ü Sorting numerically ü Factoring big numbers ü Playing chess ü Playing poker ü Playing go ü Face recognition ü
  4. 7 Cognitive Task Human Machine (2018) Adding 4-digit numbers ü

    Adding 5-digit numbers ü ... ü Adding 8923-digit numbers ü Spelling ü Sorting alphabetically ü Sorting numerically ü Factoring big numbers ü Playing chess ü Playing poker ü Playing go ü Face recognition ü Leading Pavilion seminar ?
  5. 14

  6. More Ambition 15 “The human race will have a new

    kind of instrument which will increase the power of the mind much more than optical lenses strengthen the eyes and which will be as far superior to microscopes or telescopes as reason is superior to sight.”
  7. More Ambition 16 “The human race will have a new

    kind of instrument which will increase the power of the mind much more than optical lenses strengthen the eyes and which will be as far superior to microscopes or telescopes as reason is superior to sight.” Gottfried Wilhelm Leibniz (1679)
  8. 17 Gottfried Wilhelm Leibniz (Universitat Altdorf, 1666) who advised: Jacob

    Bernoulli (Universitdt Basel, 1684) who advised: Johann Bernoulli (Universitdt Basel, 1694) who advised: Leonhard Euler (Universitat Basel, 1726) who advised: Joseph Louis Lagrange who advised: Simeon Denis Poisson who advised: Michel Chasles (Ecole Polytechnique, 1814) who advised: H. A. (Hubert Anson) Newton (Yale, 1850) who advised: E. H. Moore (Yale, 1885) who advised: Oswald Veblen (U. of Chicago, 1903) who advised: Philip Franklin (Princeton 1921) who advised: Alan Perlis (MIT Math PhD 1950) who advised: Jerry Feldman (CMU Math 1966) who advised: Jim Horning (Stanford CS PhD 1969) who advised: John Guttag (U. of Toronto CS PhD 1975) who advised: David Evans (MIT CS PhD 2000) my academic great- great-great-great- great-great-great- great-great-great- great-great-great- great-great- grandparent!
  9. More Precision 18 “The human race will have a new

    kind of instrument which will increase the power of the mind much more than optical lenses strengthen the eyes and which will be as far superior to microscopes or telescopes as reason is superior to sight.” Gottfried Wilhelm Leibniz (1679) Normal computing amplifies (quadrillions of times faster) and aggregates (enables millions of humans to work together) human cognitive abilities; AI goes beyond what humans can do.
  10. 20 The history of computer chess is the history of

    artificial intelligence. After their disappointments in trying to reverse- engineer the brain, computer scientists narrowed their sights. Abandoning their pursuit of human-like intelligence, they began to concentrate on accomplishing sophisticated, but limited, analytical tasks by capitalizing on the inhuman speed of the modern computer’s calculations. This less ambitious but more pragmatic approach has paid off in areas ranging from medical diagnosis to self-driving cars. Computers are replicating the results of human thought without replicating thought itself. Nicolas Carr, A Brutal Intelligence: AI, Chess, and the Human Mind, 2017
  11. 21

  12. Operational Definition “Artificial Intelligence” means making computers do things their

    programmers don’t understand well enough to program explicitly. 23 If it is explainable, its not AI! Note: you will definitely have opportunities to argue for an alternative definition.
  13. Promise of AI 24 Automate tasks that are dangerous, tedious,

    unenjoyable, etc. for humans Reduce the costs of design, production, assembly, distribution for all products to nearly 0 Mistake-free, continually-improving, low-cost medical care for all Optimal decision-making systems with complete knowledge and no human cognitive biases
  14. Potential Harmful Impacts of AI Benign developers and operators AI

    out of control AI inadvertently causes harm Malicious operators Build AI to do harm 25
  15. Harmful AI Benign developers and operators AI out of control

    AI causes harm (without creators objecting) Malicious operators Build AI to do harm 26
  16. Harmful AI Benign developers and operators AI out of control

    AI inadvertently causes harm to humanity Malicious operators Build AI to do harm 29
  17. Harmful AI Benign developers AI out of control AI causes

    harm (without creators objecting) Malicious developers Using AI to do harm 32 Malice is (often) in the eye of the beholder (e.g., mass surveillance, pop-up ads, etc.)
  18. 33 “The future has arrived — it’s just not evenly

    distributed yet.” (William Gibson, 1990s) Photo: Christopher J. Morris/Corbis
  19. 34 “The future has arrived — it’s just not evenly

    distributed yet.” (William Gibson, 1990s) Expanding victims: Attacks that are only cost-effective for high-value, easy-compromise targets, become cost-effective against everyone Expanding adversaries: Attacks only available to nation-state level adversaries, become accessible to everyone
  20. Automated Spear Phishing 36 “It’s slightly less effective [than manually

    generated] but it’s dramatically more efficient” (John Seymour)
  21. Potential Harmful Impacts of AI Benign developers and operators AI

    out of control AI inadvertently causes harm Malicious operators Build AI to do harm 39 Main goal of this seminar is to better understand these potential harms, and possible ways to mitigate them.
  22. Majors in the Class 41 15 Students 14 Different Majors

    (of 24 total) 7 CompSci 4 CogSci 2 Economics, Government 1 of the others 7 - 3rd years 8 - 4th years
  23. Aspirations 42 What you wanted to be when you grew

    up when you were seven Engineer or an Astronomer A chef a biomedical engineer that builds prosthetics Lawyer a dolphin trainer I don't think I had any idea. I honestly can't remember but I did a project in second grade on Amelia Earhart so it might've been a pilot. a spy or journalist I wanted to be a spy/ninja. I thought I wanted to be a pharmacist, which was mostly because even at age 7 I liked the medical field, but preferred the hours of a pharmacist over those of a physician. When I was 7 years old I deeply wanted to be a doctor. Mostly because I hated going to the doctor and thought I could do a better job. This dream quickly died away when I realized how squeamish I am around blood. An astronaut, and then an aeronautical engineer after I heard about the NASA height restrictions A teacher A singer
  24. Aspirations 43 What you want to be when you “grow

    up” now entrepreneurial: “social entrepreneur using technology”, “run purpose-driven tech company”, “my own boss” creator: create films or score films/video games; video content creator, digital storyteller policy: lawyer for an advocacy group, “work in the technology sector to shape a better future”, “oversee projects that deal with technological advances” science: aviation medicine researcher (neuropsychological issues in military personnel), paleontologist “commercial airline pilot or work in the mental health field”, “something cybersecurity related”, “human-computer interaction” personal: “want to be a really good dad (eventually)”, “when I grow up, I want to be learning”
  25. Should we be worried? 46 Cal, Mira Kavya, Olivia Megan,

    Stella Lauren (A.), Nathaniel Lauren (H.), Rohit Emily, Erich Fiona, Jacob, Maria
  26. 47 More than 10,000 years ago, 1 100-1000 years ago,

    1 50-100 years ago, 1 20-50 years ago, 2 5-20 years ago, 3 Present day, 3 5-20 years from now, 2 50-100 years from now, 2 Matched to maximize group disparity
  27. 48 On Robots Joe Berger and Pascal Wyse (The Guardian,

    21 July 2018) Human Jobs of the Future
  28. 49 Paradoxically, a series of ‘improvements’, each of which was

    meant to make life easier, added up to a millstone around the necks of these farmers. Why did people make such a fateful miscalculation? For the same reason that people throughout history have miscalculated. People were unable to fathom the full consequences of their decisions.
  29. 52 “From the beginning of civilization until the industrial revolution

    a man could, as a rule, produce by hard work little more than was required for the subsistence of himself and his family, although his wife worked at least as hard and his children added their labor as soon as they were old enough to do so. The small surplus above bare necessaries was not left to those who produced it, but was appropriated by priests and warriors.” Bertrand Russell (1872-1970) In Praise of Idleness 1932
  30. 53 “In a world where no one is compelled to

    work more than four hours a day, every person possessed of scientific curiosity will be able to indulge it, and every painter will be able to paint without starving, ..., teachers will not be exasperatedly struggling to teach by routine methods things which they learnt in their youth, which may, in the interval, have been proved untrue..” Bertrand Russell (1872-1970) In Praise of Idleness 1932
  31. 54 “Modern methods of production have given us the possibility

    of ease and security for all; we have chosen instead to have overwork for some and starvation for others. Hitherto we have continued to be as energetic as we were before there were machines. In this we have been foolish, but there is no reason to go on being foolish for ever.” Bertrand Russell (1872-1970) In Praise of Idleness 1932
  32. Seminar Structure - Other meetings won’t be like this one!

    - You’ll be arriving having read something to discuss - You’ll have prepared responses to some questions about the readings - What we do will largely be guided by students - Make suggestions! I am open to (almost) anything - If you don’t like the assigned reading, propose an alternative - Not everyone needs to do the same thing 56
  33. Main Assignments Two main “papers” Satisfy Second Writing Requirement: 58

    Empty Word document is 13,594 bytes = 3398.5 (32-bit) words
  34. Main Assignments Two main “papers” Satisfy Second Writing Requirement: 59

    Don’t write for length: write to communicate clearly, convincingly, and concisely Papers should be substantial: - tackle a challenging topic - original and interesting ideas - enough substance that it wouldn’t fit into a short paper
  35. What I mean by “paper” Something that involves writing in

    English essay, blog post, scripted video, Jupyter notebook to communicate an original message some important, new idea using evidence and argument facts, sources, logic to a target audience. not just me, not just your classmates, hopefully wider posted publicly and permanently! 60
  36. Weekly Assignments - Readings (sometimes viewing) posted on course site

    - Reactions to readings Fact check at least one claim Responses (can be based on provided questions) - Contributions to Class Forum Post links to articles with comments Comment on things other post 61
  37. 63 "The hardcover American edition of Sapiens weighs two and

    a half pounds—a little less than the average weight of a Homo sapiens brain. This is unusual for something that is neither a reference work nor a coffee-table book, and that runs to fewer than five hundred pages. The reason for such disproportionate heft is the quality of the paper: the pages are thick like those of a book of prints, crisp white and replete with color illustrations.” John Sexton, A Reductionist History of Humankind
  38. Not Everyone Loves It 66 “Summing up the book as

    a whole, one has often had to point out how surprisingly little he seems to have read on quite a number of essential topics. It would be fair to say that whenever his facts are broadly correct they are not new, and whenever he tries to strike out on his own he often gets things wrong, sometimes seriously. So we should not judge Sapiens as a serious contribution to knowledge but as 'infotainment', a publishing event to titillate its readers by a wild intellectual ride across the landscape of history, dotted with sensational displays of speculation, and ending with blood-curdling predictions about human destiny. By these criteria it is a most successful book.” C. R. Hallpike's review
  39. Not Everyone Loves It 67 “The book is fundamentally unserious

    and undeserving of the wide acclaim and attention it has been receiving. But it is worth considering the book’s blind spots and flaws—the better to understand the weaknesses of the genre and the intellectual temptations of our age.” John Sexton, A Reductionist History of Humankind
  40. Before Next Meeting Reading (see Week 1 on course site):

    Sapiens, Chapter 1-8 Why Technology Favors Tyranny Reactions: Fact check at least one claim Responses (questions provided) Class Forum Post links to article with comment Comment on things other post 69