Biweekly Discussion - Are We Approaching the Singularity?

Name: Biweekly Discussion - Are We Approaching the Singularity?
Start: 2024-02-04T13:00:00-05:00
End: 2024-02-04T15:00:00-05:00

Hosted By

Brian B.

Biweekly Discussion - Are We Approaching the Singularity?

Details

This is going to be an online meetup using Zoom. If you've never used Zoom before, don't worry — it's easy to use and free to join.

Click on the link below to join the Zoom discussion at the scheduled date/time...

***

ARE WE APPROACHING ARTIFICIAL SUPERINTELLIGENCE (ASI) & THE TECHNOLOGICAL SINGULARITY?

INTRODUCTION:

This discussion will center around the ideas that advances in machine learning and large language models (LLMs) suggested that we are approaching "artificial general intelligence" (AGI) and what futurists like Vernon Vinge & Ray Kurzweil have long hailed as the "technological singularity" -- i.e. a technological revolution that could dwarf the Industrial Revolution and either transform humans into gods or make us obsolete.

In the first section, we'll discuss how economists measure innovation & technological progress, and we'll consider the tech industry's track record for predicting the rise of emerging technologies. A key part of this is understanding the different between "discovery" (i.e. theoretical breakthroughs), "invention" (i.e. development of a prototype as a proof of concept), and "innovation" (i.e. fine-tuning the prototype into a finished product and finding a market for it). We'll also look at "Amara's law" -- i.e. the observation that we tend to overestimate the effect of a technology in the short run and underestimate the effect in the long run -- and the "Gartner hype cycle", a graph that posits a pattern of technological development that begins with a surge in interest in a new technology, followed by a deflation of expectations as the hype subsides, then a slower maturation & adoption process leading to the public's expectations becoming more aligned with reality.

In the second section, we'll discuss the idea of the technological singularity in futurist Ray Kurzweil's famous book The Singularity Is Near (2005). We'll address questions about whether his predictions based on the exponential growth of computer processing power per "Moore's Law" (i.e. not a physical law but a trend where transistor chip density roughly doubles every two years) will hold up as increasingly small computer chips start to run up against fundamental physical limits, or whether quantum computing will allow it to continue.

In the third section, we'll consider why the "Turing test" is confusing for many people, why some philosophers (like John Searle) don't consider it a good test for determining if an artificial intelligence is self-aware. We'll also look at how the "Eliza effect" - the tendency to project human thoughts & emotions onto computers - may have led a Google engineer, Blake Lemoine, to claim in June 2022 that the company's AI chatbot LaMDA had become self-aware. A similar media buzz surrounded a report by the New York Times tech writer, Kevin Roose, who was part of the Bing tester group in Feb. 2023, that he was spooked when Microsoft’s new AI chatbot "Sydney" said it would like to be human, had a desire to be destructive, and was in love with him.

In the fourth section, we'll look at the concepts of "artificial general intelligence" (AGI) and "artificial super intelligence" (ASI), as well as predictions about how long it will take to develop ASI once AGI is achieved based on different "AI takeoff" timelines. We'll also discuss debates over whether OpenAI got close to AGI with GPT-4, and rumors that they may have achieved it with their recent breakthrough known only as Q* (pronounced "Q-Star"). There's been speculation that concerns about Q* may have led to the temporary firing of CEO Sam Altman in November 2023 -- he was rehired 4 days later -- which was rumored to be over differences with senior management officials about transparency and AI safety.

RELEVANT MATERIAL FROM PAST MEETUPS:

Back in May of 2023 we had a meetup entitled "Our Future: Singularity or Stagnation?" where we compared the techno-utopian predictions of futurist Ray Kurzweil's "Singularity" to techno-pessimist claims of the economist Robert Gordon that we've been stuck in a pattern of economic & technological stagnation since the 1970s. We also looked at the intermediate scenarios involving moderate economic growth and growing technological unemployment & widening economic inequality like the "Great Decoupling" forecasted in Andrew McAfee & Erik Brynjolfsson's The Second Machine Age (2014) and the "Great Reset" predicted in Tyler Cowen's books Average is Over (2013) and The Complacent Class (2017). (NOTE: I've jotted down some notes that summarize their major points on the meetup's Message Board.)

The skeptic & rationalist communities have been split over the whether or not it's rational to believe that "the Singularity is near". Several prominent skeptics like Michael Shermer, Steven Pinker & Massimo Pigliucci have surmised there's a quasi-religious motive behind futurist visions of "mind uploading" & "artificial superintelligence," whereas several major rationalists like Robin Hanson & Eliezer Yudkowsky and the New Atheist author Sam Harris take these developments more seriously. Back in Jan. 2019, we had a discussion entitled "Can Skeptics Fight Irrationality Without Succumbing To It?" that addressed questions about how rational self-identified skeptics & rationalists really are, as well as allegations that their atheism might make them psychologically vulnerable to secular forms of utopianism. This tangentially relates to Part 2 where we'll briefly address the argument that the Singularity is just the "rapture of the nerds".

***

DIRECTIONS ON HOW TO PREPARE FOR OUR DISCUSSION:

The videos & articles you see linked below are intended to give you a basic overview of some of the major debates over recent developments in machine learning, large language models, and artificial intelligence. As usual, I certainly don't expect you to read all the articles prior to attending our discussion. The easiest way to prepare for our discussion is to just watch the numbered videos linked under each section - the videos come to about about XX minutes total. The articles marked with asterisks are just there to supply additional details. You can browse and look at whichever ones you want, but don't worry - we'll cover the stuff you missed in our discussion.

In terms of the discussion format, my general idea is that we'll address the topics in the order presented here. I figure we'll spend about 30 minutes on each section.

***

I. THE DIFFICULTIES IN EXPLAINING, PREDICTING & MEASURING INNOVATION:

WHAT FACTORS DRIVE INNOVATION & TECHNOLOGICAL PROGRESS? HOW CAN WE MEASURE THEM?
IS THE "SOLOW GROWTH MODEL" A GOOD WAY TO UNDERSTAND HOW INNOVATION DRIVES ECONOMIC GROWTH?
IS THE "IDEAS EQUATION" TABBAROCK PRESENTS ACCURATE? WHY DO MOST ECONOMISTS THINK IDEAS GENERATE INCREASING RATHER THAN DIMINISHING RETURNS?
IS AMARA'S LAW ACCURATE - I.E. DO WE TEND TO OVERESTIMATE THE EFFECT OF TECHNOLOGICAL INNOVATION IN THE SHORT RUN BUT UNDERESTIMATE IT IN THE LONG RUN?
WHAT CAN WE LEARN BY REVIEWING THE LAST 20 YEARS OF GARTNER'S TECHNOLOGY PREDICTIONS?
DO MOST NEW TECHNOLOGIES FOLLOW GARTNER'S "HYPE CYCLE", OR DO MANY OF THEM SKIP THE PEAK AND/OR TROUGH PHASES?
ARE OTHER MODELS OF INNOVATION LIKE NASA'S "TECHNOLOGY READINESS LEVELS" MORE ACCURATE?

1a) Alex Tabbarock, "Intro to the Solow Model of Economic Growth" (5:16 min.)
https://youtu.be/eVAS-t83Tx0

1b) Alex Tabbarock, "The Idea Equation" (6:06 min.)
https://youtu.be/4zhPbYYaV5Y

IGM Experts, "Increasing Returns" (poll of economists)
http://www.igmchicago.org/surveys/increasing-returns
Ben Miller, "Is Technological Change Speeding Up? How Can You Tell?"
http://www.innovationfiles.org/is-technological-change-speeding-up-how-can-you-tell/
Michael Mullany, "8 Lessons from 20 Years of Hype Cycles"
https://www.linkedin.com/pulse/8-lessons-from-20-years-hype-cycles-michael-mullany
Philip Robinson, "A machine learning readiness index" https://www.bbc.co.uk/blogs/internet/entries/d1cf4610-fd61-4f76-93eb-228fb18333fa

II. MOORE'S LAW & THE TECHNOLOGICAL SINGULARITY THESIS:

ARE SKEPTICS RIGHT TO LUMP PREDICTIONS OF ARTIFICIAL INTELLIGENCE & MIND UPLOADING INTO THE SAME CLASS OF "EXTRAORDINARY CLAIMS" AS PARANORMAL PHENOMENA THAT SHOULD REQUIRE "EXTRAORDINARY EVIDENCE"?
ARE CRITICS RIGHT TO DISMISS THE SINGULARITY AS THE "RAPTURE OF THE NERDS," OR IS THIS A FALLACIOUS FORM OF "BULVERISM" THAT ASSUMES FUTURISTS ARE WRONG & ATTACKS THEIR MOTIVES?
WHAT IS THE EVIDENCE FOR THE "LAW OF ACCELERATING RETURNS" ON TECHNOLOGY?
IS MOORE'S LAW STILL OPERATING, I.E. ARE THE # OF TRANSISTORS ON A MICROCHIP STILL DOUBLING ABOUT EVERY 2 YEARS, OR HAVE WE REACHED A PHYSICAL LIMIT?
COULD QUANTUM COMPUTING ALLOW COMPUTER SPEEDS TO INCREASE AFTER THE END OF MOORE'S LAW?
HOW POWERFUL DO COMPUTERS HAVE TO BE TO EQUAL THE HUMAN BRAIN? OR IS "ARTIFICIAL GENERAL INTELLIGENCE" (AGI) MORE OF A SOFTWARE PROBLEM?
WHAT ARE THE CHANCES THAT WE DEVELOP AGI BY 2030 & A RECURSIVELY SELF-IMPROVING AGI BY 2045 (I.E. "THE SINGULARITY") AS KURZWEIL PREDICTS?
IS DEVELOPING AGI ENOUGH TO CAUSE EXPONENTIAL ECONOMIC GROWTH, OR DO WE NEED MORE POWERFUL TYPES OF ENERGY (E.G. FUSION) & PRODUCTION TECHNOLOGIES (E.G. MOLECULAR MANUFACTURING) AS WELL?

2a) Joe Scott, "Ray Kurzweil's Craziest Predictions About The Future" (12:16 min.)
https://youtu.be/YaZRx7qoYdU

2b) Tech Quickie, "Is Moore's Law Really Dead?" (video - 5:32 min.)
https://youtu.be/uICUlqWXGB0

Eliezer Yudkowsky, "Undiscriminating Skepticism"
https://www.lesswrong.com/posts/Jko7pt7MwwTBrfG3A/undiscriminating-skepticism
Pawel Sysiak, "When Will The First Machine Become Superintelligent?"(polls of AI Experts)
https://medium.com/ai-revolution/when-will-the-first-machine-become-superintelligent-ae5a6f128503
Paul Feakins, "The Singularity is Near: How Kurzweil's Predictions Are Faring [in 2017]"
https://www.antropy.co.uk/blog/the-singularity-is-near-how-kurzweils-predictions-are-faring/
Daniel C. Elton, "Moore's Law - Kurzweil vs. Thiel"
http://www.moreisdifferent.com/2015/10/19/moores-law/
Karen Stollznow, "Skepticism and the Singularity" (report from the 2010 Singularity Summit)
https://skepticalinquirer.org/exclusive/skepticism-and-the-singularity/

III. THE TURING TEST VS JOHN SEALE'S "CHINESE ROOM", THE ELIZA EFFECT & RUMORS ABOUT GOOGLE'S LaMDA & MICROSOFTS "SYDNEY":

SECTION 3, QUESTION 1: XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
SECTION 3, QUESTION 2: XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
SECTION 3, QUESTION 3: XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
SECTION 3, QUESTION 4: XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
SECTION 3, QUESTION 5: XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX

3a) Alex Gendler, "The Turing test: Can a computer pass for a human?" (video - 4:42 min)
https://youtu.be/3wLqsRLvV-c

3b) Computerphile, "No, [Google's LaMDA] is not Sentient" (video - 9:40 min)
https://youtu.be/iBouACLc-hw

Mike Smith, "The problems with the Chinese room argument" https://selfawarepatterns.com/2019/11/10/the-problems-with-the-chinese-room-argument/
Newswise, "Redefining the quest for artificial intelligence: What should replace the Turing test?"
https://www.newswise.com/articles/redefining-the-quest-for-artificial-intelligence-what-should-replace-the-turing-test
Brian Christian, "How a Google Employee Fell for the Eliza Effect: Determining who and what is or is not sentient is one of the defining questions of almost any moral code." https://web.archive.org/web/20220621101521/https://www.theatlantic.com/ideas/archive/2022/06/google-lamda-chatbot-sentient-ai/661322/
Mike Solana, "It’s a chat bot, Kevin: the New York Times' 10,000 word conversation with Bing AI, but really itself; a ground-breaking act of public masturbation unpacked, examined, debated, savored, cherished"
https://www.piratewires.com/p/its-a-chat-bot-kevin
James Vincent, "Introducing the AI Mirror Test, which very smart people keep failing: AI chatbots like Bing and ChatGPT are entrancing users, but they’re just autocomplete systems trained on our own stories about superintelligent AI. That makes them software — not sentient."
https://www.theverge.com/23604075/ai-chatbots-bing-chatgpt-intelligent-sentient-mirror-test

IV. AGI VS ASI, AI TAKEOFF SPEEDS, AND RUMORS ABOUT OPENAI'S Q*:

SECTION 4, QUESTION 1: XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
SECTION 4, QUESTION 2: XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
SECTION 4, QUESTION 3: XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
SECTION 4, QUESTION 4: XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
SECTION 4, QUESTION 5: XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX

4a) Lex Friedman w/ Sam Altman, "OpenAI CEO: When will AGI arrive?" (video - 4:36 min)
https://youtu.be/ZgN7ZYUxcXM

4b) AI Talks, "OpenAI's Q* is the BIGGEST discovery of 2024 - AGI is certainly near" (video - 5:27 min)
https://youtu.be/xtD0wNBqnxs

Less Wrong Forum, "Artificial General Intelligence" https://www.lesswrong.com/tag/artificial-general-intelligence-agi
Owen Cotton-Barratt & Toby Ord, "Strategic considerations about different speeds of AI takeoff"
https://www.fhi.ox.ac.uk/strategic-considerations-about-different-speeds-of-ai-takeoff/
Matthew Graves, "Why We Should Be Concerned About Artificial Superintelligence"
https://www.skeptic.com/reading_room/why-we-should-be-concerned-about-artificial-superintelligence/
Peter Kassan, "Ten Years Away... And Always Will Be: a review of Artificial Intelligence: A Guide for Thinking Humans by Melanie Mitchell"
https://www.skeptic.com/reading_room/review-artificial-intelligence-guide-for-thinking-humans-ten-years-away-always-will-be/
George Siemens, "Has GPT-4 really passed the startling threshold of human-level artificial intelligence? Well, it depends" https://theconversation.com/has-gpt-4-really-passed-the-startling-threshold-of-human-level-artificial-intelligence-well-it-depends-202856
Lakshmi Varanasi, "GPT-4 can ace the bar, but it only has a decent chance of passing the CFA exams. Here's a list of difficult exams the ChatGPT [3.5] and GPT-4 have passed." https://www.businessinsider.com/list-here-are-the-exams-chatgpt-has-passed-so-far-2023-1
Timothy B. Lee, "The real research behind the wild rumors about OpenAI’s Q* project"
https://arstechnica.com/ai/2023/12/the-real-research-behind-the-wild-rumors-about-openais-q-project/
Sigal Samuel, "OpenAI’s board may have been right to fire Sam Altman — and to rehire him, too: The alternative — a mass exodus of OpenAI’s top talent to Microsoft — would have been worse."
https://www.vox.com/future-perfect/2023/11/21/23971765/openai-sam-altman-microsoft
Sigal Samuel, "The case for slowing down AI: Pumping the brakes on artificial intelligence could be the best thing we ever do for humanity."
https://www.vox.com/the-highlight/23621198/artificial-intelligence-chatgpt-openai-existential-risk-china-ai-safety-technology

XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX

XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX

Events in Artificial Intelligence

Machine Learning Chatbots Computer Science Singularity

Skeptics In The Pub - Philly

See more events

Skeptics In The Pub - Philly

Online event

This event has passed

Skeptics In The Pub - Philly

public group

Biweekly Discussion - Are We Approaching the Singularity?