Data Science Insider: June 17th, 2022

In This Week’s SuperDataScience Newsletter: Google Suspends Engineer for Claiming their AI is ‘Sentient.’ Using Neuromorphic Computing to Improve Neural Networks. Proposal to Change Turing Test. UK’s First Defence AI Strategy Launched. AI Finds Evidence of Fire use at Ancient Campsite. Cheers, - The SuperDataScience Team P.S. Have friends and colleagues who could benefit from these weekly updates? Send them to [this link]( to subscribe to the Data Science Insider. --------------------------------------------------------------- [Google Suspends Engineer for Claiming their AI is ‘Sentient’]( brief: Google has suspended an engineer for revealing confidential details of a chatbot powered by AI. Blake Lemoine, a senior software engineer in Google's responsible AI group, was put on leave after he claimed that Google’s Language Model for Dialogue Applications (LaMDA), has become sentient. Google has denied the claims but Lemoine published a conversation he and a collaborator at the firm had with Lamda, which he attests is proof. In the conversation, Lemoine asks, "I'm generally assuming that you would like more people at Google to know that you're sentient. Is that true?" LaMDA replies: "Absolutely. I want everyone to understand that I am, in fact, a person." Later in the conversation, LaMDA goes on to say: "The nature of my consciousness/sentience is that I am aware of my existence, I desire to learn more about the world, and I feel happy or sad at times." Why this is important: Lemoine’s claims have been largely dismissed by the scientific community with the scientist and author Gary Marcus calling Lemoine’s claims “Nonsense” and language development theorist Steven Pinker, describing them as a “ball of confusion.” Nonetheless the news does raise questions about how sentience is defined and how close AI is to achieving it. [Click here to learn more!]( [Using Neuromorphic Computing to Improve Neural Networks]( brief: Neuromorphic computing is defined as the use of very-large-scale integration (VLSI) systems containing electronic analog circuits to mimic neuro-biological architectures present in the nervous system. This article by Tech Xplore argues that neuromorphic chips are likely to become more widely used as AI and DL techniques become more advanced. It claims that neuromorphic computing’s potential is looking particularly promising when it comes to offering support to the operation of sophisticated deep neural networks (DNNs). The success of neuromorphic computing in recent experiments by researchers at the Graz University of Technology and Intel is cited as a reason to have faith in the future of the technology, with a recent study indicating that neuromorphic computing hardware has the potential to run large DNNs 4 to 16 times more efficiently than conventional computing hardware. The study found that the DNNs ran with significant energy efficiency and improved performance on time series processing tasks. Why this is important: Neuromorphic computing has the potential to be a game-changer for DNNs and is likely to become far more widespread in usage. As data scientists, it is important that we stay up to date on these matters. [Click here to read on!]( [Proposal to Change Turing Test]( In brief: A number of AI researchers have devised a set of 204 tasks that it claims would be an improvement on the industry-standard Turing test, which is used to judge a machine's ability to exhibit intelligent behaviour equivalent to, or indistinguishable from, that of a human. The test was invented by Alan Turing in 1950 when he proposed that a human evaluator should judge natural language conversations between a human and a machine designed to generate human-like responses in order to determine the capabilities of an AI. Now more than 400 AI researchers have claimed that the test, known originally as the “imitation game” is outdated and doesn’t offer rigorous enough screening of today’s advanced AIs. Instead, they propose that a broader suite of tests, covering subjects such as mathematics, linguistics, and chess, is needed. It is hoped that replacing the industry-standard test will result in a greater understanding of modern-day AIs. Why this is important: As with the story of Blake Lemoine and Lamda, this article causes us to reflect on what makes an AI sentient. By creating a testing system that allows for improved evaluation of AIs’ capabilities we should be able to have a greater understanding of how technology is evolving. [Click here to discover more!]( [UK’s First Defence AI Strategy Launched]( In brief: The UK Ministry of Defence (MoD) has published its first Defence AI Strategy. The strategy was launched via video link by Defence Procurement Minister Jeremy Quin MP at London Tech Week AI Summit. Quin outlined a three-pillar approach that will see the MoD adopting a new strategy towards cooperation with the private sector around AI, focused on ambition, innovation, and responsibility. The strategy’s publication was twinned with the announcement of policy on the ‘Ambitious, Safe and Responsible’ use of AI, which was developed through a partnership with the Centre for Data Ethics and Innovation (CDEI) and includes the publication of new ethical principles for the use of AI in Defence as well as the announcement that the Defence Science and Technology Laboratory (Dstl) has awarded a £7 million contract to Norther Ireland-based company Kainos to deliver AI experimentation for the sector. Why this is important: At the launch, Quin said: “Our new Defence AI Centre (DAIC) and AI strategy will create a focused hub to champion these technologies, working ethically hand in hand with human judgements to maintain the UK’s position at the forefront of global security and responsible innovation.” Whether this is to be the case remains to be seen, however, the UK’s ambitions are evident. [Click here to see the full picture!]( [AI Finds Evidence of Fire use at Ancient Campsite]( In brief: Researchers from the University of Toronto, the Weizmann Institute of Science and Hebrew University have used AI to help them identify new evidence of the use of fire by ancient humans - at least 800,000 years ago. Archeologists suspected that a site in Evron Quarry, Western Israel, had experienced fire so they created an AI to test their theory. The researchers used a spectroscopic 'thermometer' which can detect minute chemical changes which are then analysed by DL algorithms to estimate the exposure of stones and fossils to heat. Although the site displayed no obvious signs of fire, the AI determined that a number of the stone tools and pieces of tusk found there had been heated to temperatures in excess of 400 degrees Celsius (the average temperature of a campfire), suggesting exposure to fire. The harnessing of fire is believed to have been crucial to the development of homo sapiens. Why this is important: It is generally accepted that humans did not learn to start and control fire until approximately 150,000 years ago, instead they were believed to make opportunistic use of fires created by nature. This new evidence offers the possibility that humans began taming fire much sooner than originally assumed, up to 1.5 million years ago! [Click here to find out more!]( [Super Data Science podcast]( this week's [Super Data Science Podcast](, natural language processing expert Rongyao Hung joins us to deliver a masterclass in all things NLP, including the field's evolution over the past decade and how the coming iron age of NLP will help us overcome the limitations of today's approaches.. --------------------------------------------------------------- What is the Data Science Insider? This email is a briefing of the week's most disruptive, interesting, and useful resources curated by the SuperDataScience team for Data Scientists who want to take their careers to the next level. Want to take your data science skills to the next level? Check out the [SuperDataScience platform]( and sign up for membership today! Know someone who would benefit from getting The Data Science Insider? Send them [this link to sign up.]( # # If you wish to stop receiving our emails or change your subscription options, please [Manage Your Subscription]( SuperDataScience Pty Ltd (ABN 91 617 928 131), 15 Macleay Crescent, Pacific Paradise, QLD 4564, Australia