Artificial intelligence author kicks off Friends of the Library nonfiction lecture series – Naples Daily News

Over the past few decades, a bunch of smart guys built artificial intelligence systems that have had deep impact on our everyday lives. But do they — and their billion-dollar companies — have the human intelligence to keep artificial intelligence safe and ethical?
Questions like this are part of the history and overview of artificial intelligence in Cade Metz’s book “Genius Makers: The Mavericks Who Brought AI to Google, Facebook, and the World.”
On Monday, Jan. 17, Metz, a technology correspondent for The New York Times and former senior writer for Wired magazine, is the first speaker in the 2022 Nonfiction Author Series, sponsored by the nonprofit Friends of the Library of Collier County, which raises money for public library programs and resources.
More about the series: Collier County Nonfiction Author Series set for early 2022, will be in-person
AND: Florida-based authors headline Friends of the Bonita Springs Library luncheon
The lecture series includes breakfast and is being held this year at a new venue, the Kensington Country Club in Naples. The series is sold out, but you can contact the Friends to be put on a waiting list. (See info box for details.)
Metz grew up in Raleigh, North Carolina, where his parents met while both working for IBM, so computing is in his blood. His father helped develop the Universal Product Code (UPC) — that ubiquitous bar code that now is on absolutely everything. Metz attended Duke University as an IBM scholarship student, majoring in English and planning to be a writer, while also working at IBM as a programmer.
“Genius Makers,” his first book, centers on advances in technology, but his real impetus was to write about the fascinating characters who were developing these ideas and visions. The book focuses on two unusual men whose research in artificial intelligence has driven a technology arms race. And it raises intriguing questions, such as: What does it mean to be human?
Metz answered some questions ahead of his talk in Naples.
Naples Daily News: What are the most common, everyday example of how AI (artificial intelligence) has affected the world in the past 20 years?
Cade Metz: The best examples are talking digital assistants like Siri and Alexa, which have improved significantly over the past decade. They can recognize spoken words with the accuracy of a human. Their synthetic voices are increasingly lifelike. And though they have not yet reached the point where they can actually carry on a conversation — really understand the meaning of what they are hearing and properly respond to it — their language skills continue to improve.
Meanwhile, the fundamental concepts that underpin these digital assistants are driving a wide range of other technologies, including online services like Google Translate that instantly translate between languages and warehouse robots that sort through giant bins of random stuff.
NDN: The dream of self-driving cars is, for most people, the face of how AI could change our lives. How realistic do you think a true, safe self-driving car is in, say, the next decade?
CM: This technology continues to improve. But it is still a long way from everyday life. Only one company — a Google spinoff called Waymo — is actually offering a self-driving car service, and that is in the suburbs of Phoenix, Arizona, where the roads are wide, pedestrians are few and the weather is good. When it rains, the company halts the service, and at times, when the cars are unable to navigate on their own, the company uses remote control software to get them going again. What this means is that it will likely be a decade or more before these vehicles are commonplace.
NDN: This was such a wonderful sentence early on in your book: “As an undergraduate at Harvard (in the 1940s), using over three thousand vacuum tubes and a few parts from an old B-52 bomber, (Marvin) Minsky built what may have been the first neural network.” Is that kind of amateur, garage-built science still possible, given the speed of innovation now and the billions of dollars that are thrown at development?
CM: It certainly is. It happens all the time, inside universities and out. But in the AI field, this has been eclipsed by the work at giant companies like Google and Facebook. That is one of the major threads in my book: academia struggling to keep up with the rapid rate of progress in the tech industry. It is a real problem. So much of the talent is moving into industry, leaving the cupboard bare at universities. Who will teach the next generation? Who will keep the big tech companies in check? 
NDN: I was amused to see that Google and DeepMind built a team “dedicated to what they called ‘AI safety,’ an effort to ensure that the lab’s technologies did no harm.” My question is, who defines harm within this race to monetize new technologies? Isn’t, for example, the staggering amount of electrical power used to run these systems harmful to the globe?
CM: I am glad you were amused. These companies say we should trust them to ensure AI “safety” and “ethics,” but the reality is that safety and ethics are in the eye of the beholder. They can shape these terms to mean whatever they like. Many of the AI researchers at the heart of my book are genuinely concerned about how AI will be misused — how it will cause harm — but when they get inside these large companies, they find that their views clash with the economic aims of these tech giants.
NDN: Along the same lines, you address how the neural networks “learn” by hoovering up data from the web. Since much of what’s on the web is false or misleading — sometimes inadvertently, sometimes on purpose — what’s the gatekeeper to ensure that what’s “learned” is accurate? Even the word “accurate” is often subjective now.
CM: A neural network — the idea at the heart of modern AI — is a mathematical system that learns tasks by analyzing data. By pinpointing patterns in thousands of cat photos, for instance, a neural network can learn to identify a cat. This is the technology that allows Siri to recognize spoken words. It lets Google Translate and Skype translate from one language to another. Trouble is that this technology learns from such enormous amounts of data, we humans can’t wrap our head around it all. The designers of these systems can’t always see the false, misleading or biased information that ends up defining the technology’s behavior.
This is a huge issue for a new kind of system that learns language skills from all sorts of text posted to the internet. The internet, of course, is filled with false and biased information — not to mention hate speech and so many other things we don’t want our machines learning from. What is and what is not biased is subjective. In today’s world, what is and what is not fake news is subjective. So, yes, who will be the gatekeeper? Google? Facebook? Government regulators? We don’t know.
NDN: Could you talk about gender and racial biases? That section of the book was fascinating, such as AI’s inability to differentiate Black faces because the network hadn’t seen enough Black people to learn.
CM: This is a very real problem. Researchers have shown that face recognition systems, speech recognition systems and the latest conversational systems can be biased against women and people of color. This is often because the technology is built by white men who don’t realize they are training these systems with data that reflects only part of our society. The good news is that tech companies are waking up to the issue, and many activists and researchers are pushing for change. But it is sometimes a hard problem to solve. And, yes, the companies often have their own view of what is and what is not biased.
The Nonfiction Author Series also has announced its 2022 sponsors. Platinum sponsors are Bigham Jewelers, John R. Wood Properties, Stock Development and The Club at Olde Cypress; Gold sponsors are Books-a-Million, Gulf Coast International Properties, Naples MacFriends User Group and The Capital Grille; Silver sponsors are Tradewind Pools and Wynn’s Market.
Before each author’s presentation, a drawing will be held among ticket holders for a $250 gift certificate from Bigham Jewelers and $100 gift card from The Capital Grille.
What: Author lectures and breakfasts that are a major fundraiser for the Collier County Public Library system
Where: Kensington Country Club, 2700 Pine Ridge Road, Naples
When: Breakfast is served at 8:30 a.m.; authors speak at 9:15 a.m., followed by a book signing
Author lineup: Cade Metz, Monday, Jan. 17; Catherine Grace Katz, Monday, Feb. 14; Jared Diamond, Monday, March 7; and Jonathan Kaufman, Monday, March 28
COVID precautions: Kensington Country Club has a protocol based on CDC guidelines.  On an honor basis, people who are sick or who have symptoms should not attend; people who are vaccinated need not wear a mask; people who are not vaccinated should wear a mask until seated at their table; and people who have been sick can attend after five days isolation if they are asymptomatic and wear a mask until seated at their table.
Cost: $250 for all four events for members of the Friends of the Library of Collier County, and $295 for nonmembers. Friends memberships begin at $30/year and provide access and discounts to other programs; sign up at collier-friends.org.
Tickets: The series is sold out but there is a waiting list. Email Marlene Haywood at mhaywood@collier-friends.org or call 239-262-8135.

source
Connect with Chris Hood, a digital strategist that can help you with AI.

Leave a Reply

Your email address will not be published. Required fields are marked *

© 2022 AI Caosuo - Proudly powered by theme Octo