December 15, 2023 in Last Word
Touring with Turing
SHARE: PRINT ARTICLE:
https://doi.org/10.1287/orms.2023.04.13
Throughout the almost 70 years that artificial intelligence (AI) has been around, we have witnessed many springs and winters. In recent years, with new machine learning techniques and now with the craze of large language models (LLMs) such as ChatGPT and Bard, we have seen a resurgence in AI – so much so that soon we may see every computer with the sticker “AI Inside” instead of “Intel Chip Inside.”
I wonder what Alan Turing, the father of the Turing test, would be thinking now if he were alive. Have we achieved what Turing thought we would accomplish 70 years later?
Certainly, we have honored Turing and his achievements over the years. The Loebner Prize, in recognition of Turing and the Turing test, was an annual AI competition to see which computer programs were judged to be the most humanlike. In addition, we have the Alan Turing Institute in the U.K., where the Turing AI Scientist Grand Challenge is to develop AI systems capable of making Nobel-quality scientific discoveries highly autonomously at a level comparable, and possibly superior, to the best human scientists by 2050 [1]. In June 2021, the Bank of England issued the new 50 pound banknote to celebrate Alan Turing (although it is no longer in circulation today).
Let’s take a quick tour with Turing to see what we uncover. In today’s vogue, let’s turn first to ChatGPT and see how it answers the question, Have we achieved what Alan Turing envisioned?
“As of my last update in September 2021, it’s challenging to definitively say whether we have achieved everything that Alan Turing envisioned …
“The idea of artificial intelligence (AI) and machine learning has made significant progress since Turing’s time. We have seen AI systems capable of performing complex tasks, such as natural language processing, image recognition, and decision-making. AI has become more integrated into our daily lives, with applications in various fields like healthcare, finance, transportation, and more.
“However, it is essential to note that AI today is far from achieving full human-level intelligence (often referred to as artificial general intelligence or AGI). While AI systems have become proficient in narrow domains, they lack the depth and breadth of human intelligence. Turing’s vision of machines being able to think, learn, and adapt like humans remains an ongoing challenge in the field of AI.
“Moreover, there are also ethical and societal considerations in AI development, as exemplified by discussions on AI ethics, bias, privacy, and potential job displacement.
“Since my knowledge is limited to September 2021, I cannot provide the most current updates on AI advancements. However, it is likely that there have been further developments and progress in the field since then. It is best to consult more recent sources for up-to-date information on the state of AI and its advancements in relation to Alan Turing’s vision.”
ChatGPT has certainly stated some valid points and observations, but let’s look beyond September 2021 to see how far we have come in potentially achieving Turing’s dream.
To realize Turing’s dream, the AI system should be able to exhibit self-awareness. That is, does the AI system have moral values, self-identity, feelings and emotions? There is a group of researchers who have been looking at “Emotional AI” in recent years. Some interesting work on empathy-based chatbots and humanlike robots has emerged. The U.N.’s AI for Good Global Summit in July 2023 had a fascinating press conference with some of the robots (Ameca, Sophia, Grace) responding to questions from the press [2]. It is quite amazing how these robots exhibit humanlike behavior, intelligence and mannerisms. To address some of the emerging issues surrounding AI, the European AI Act [3] was passed by the European Parliament on June 16, 2023, as the world’s first complete set of AI rules that looks to the future of AI.
In terms of artificial general intelligence (AGI), a Brookings commentary from July 2023 on “How Close Are We to AI That Surpasses Human Intelligence?” indicates that we are quite a ways off [4]. However, others feel we are moving rapidly in the right direction, such as Demis Hassabis, CEO of Google’s DeepMind, who states that we could be “just a few years” from AI that has human-level intelligence [5]. As we move closer to AGI, we are seeing proposals for increased AI regulation to help ensure trustworthy AI.
So, let’s return to the original question regarding whether we realized Turing’s dream. In true diplomatic fashion, the answer is yes and no. We haven’t achieved AGI yet, but we are making strong strides in that direction. Of course, this creates important issues in dealing with Responsible AI, transparency, fairness, ethics, privacy, trustworthiness and the like.
There are certainly exciting times ahead! I believe my 4- and 2-year-old grandsons will be living in quite a different world as they become working adults. Hang on!
References
Jay Liebowitz recently served as the inaugural Executive-in-Residence for Public Service at Columbia University’s Data Science Institute. His main role was to infuse data science and analytics into the U.S. federal government, with support from the Partnership for Public Service. His recent books are “Pivoting Government Through Digital Transformation” (Taylor & Francis, 2024) and “Digital Transformation for the University of the Future” (World Scientific Publishing, 2023). His newest book, due to be published in mid-2024, is titled, “Regulating Hate Speech Created by GenAI” (Taylor & Francis).
([email protected])