Data Science Insider: May 20th, 2022

In This Week’s SuperDataScience Newsletter: Human-Level AI is Nearly Here. Apple ML Executive Joins Alphabet’s DeepMind. Forbes’ AI 50 2022. AI Better Than Humans at Hiring. Data Science and Dating. 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. --------------------------------------------------------------- [Human-Level AI is Nearly Here]( brief: DeepMind has claimed that it’s on the verge of achieving human-level AI with one of their research scientists and ML professor at Oxford University, Dr. Nando de Freitas, claiming that 'the game is over'. Dr. de Freitas made the bold claim regarding DeepMind’s quest to solve the hardest challenges in the race to achieve artificial general intelligence (AGI). The claim was made following the unveiling of DeepMind’s new Gato AI, dubbed a “generalist agent”, which has shown the ability to complete a wide range of complex tasks from stacking blocks to writing poetry. Dr. de Freitas’ comments came in response to an opinion piece written in The Next Web which claimed that “humans will never achieve AGI”. Speaking on Twitter, he went as far as stating that creating an AI system is capable of rivalling human intelligence is merely a matter of scaling up. Why this is important: Achieving AGI may be the goal for many in our industry but it’s not without its critics, with many concerned that the arrival of an AGI system, capable of teaching itself and becoming exponentially smarter than humans, would be impossible to switch off. [Click here to sign up!]( [Apple ML Executive Joins Alphabet’s DeepMind]( brief: Ian Goodfellow made headlines earlier this month when he quit his role as director of Machine learning at Apple due to conflicts over the firm’s work-from-home policy. Now, according to this Bloomberg article, he has agreed to join Alphabet Inc.’s DeepMind unit. Goodfellow has previously worked in a senior capacity at Google before being hired by Apple in March 2019. At the start of May, it emerged that he left over the company’s inflexible return-to-work policy that eventually requires spending three days a week at the office. Goodfellow had a number of achievements at Apple but is best known for being credited with creating generative adversarial networks (GANs). DeepMind has had a number of success stories in recent years and is widely considered a leading AI research hub. The combination of the two may prove very fruitful. Why this is important: We don’t usually cover personnel changes here at SuperDataScience but the latest development in this story reflects both a culture shift within tech companies toward a more flexible system and is something of a coup for Alphabet as Goodfellow is known as one of the world’s foremost ML researchers. An exciting future lies in store, which we’ll be sure to cover! [Click here to read on!]( [Forbes’ AI 50 2022]( In brief: It’s that time of year again when Forbes unveils its AI 50. This is the fourth annual list is produced in partnership with Sequoia Capital and “recognizes standouts in privately-held North American companies making the most interesting and effective use of artificial intelligence technology.” This year’s top spot was taken by Deepcell, a life science company noted for pioneering AI-powered cell classification and isolation for cell biology and translational research. The top 50 was compiled by a panel of independent judges formed from experts in academia, new IPO executives, venture capital, and international technology companies, following a submission process that saw over 400 entries. The article which introduced the list noted, "Inductees reflect the booming VC interest as well as the growing variability in AI-focused startups making unique uses of existing technologies, others developing their own, and many simply enabling other companies to add AI to their business model." Why this is important: The list is a prestigious ranking that illustrates what industry experts deem to be the most impressive work currently being undertaken by private companies in North America. [Click here to discover more!]( [AI Better Than Humans at Hiring]( In brief: A new study by researchers at the Inclusion Initiative at the London School of Economics and Political Science (LSE) has found that AI is equal to or better than human beings at hiring staff, however, companies still don't trust using it in the recruitment process. The study found that AI was found to be 'fairer' as well as improving the ‘fill-rate' for open positions and is 'mostly better than humans' at improving diversity in the workplace. There were limitations to AI’s capabilities with researchers finding that it had a limited talent for predicting employee outcomes after they were hired. Dr. Dario Krpan, Assistant Professor in Behavioural Science at LSE, said: “The media typically portrays AI hiring negatively and emphasises how AI can discriminate against candidates and disadvantage them. Our analysis, however, shows that even if AI is not perfect, it is fairer and more effective than human recruiters.” Why this is important: A Garter report in 2019 showed that 37% of businesses have adopted AI for recruitment. This is rapidly rising with an increase of 270% from 2015 to 2019. As AI improves, these figures are likely to increase further, giving prospecting data scientists a valuable avenue of employment. [Click here to see the full picture!]( [Data Science and Dating]( In brief: This Wired article is a short-form version of the findings presented in Seth Stephens-Davidowitz’s Don't Trust Your Gut: Using Data to Get What You Really Want in Life, a fun and interesting read which offers insights into how data science can be used to provide real-life advice which offers vast improvements in outcomes when compared to making decisions based on gut instinct alone. Stephens-Davidowitz is an economist, former Google data scientist, and New York Times bestselling author who relays the research of scientist Samantha Joel, who compiled datasets from many small studies in order to create a large enough dataset to reliably find out what predicts relationship success and what does not. Joel attempted to answer the question: “Can we help people pick better romantic partners?” and found that, ultimately, data science was unable to make accurate predictions, with relationship success proving to be unpredictable. Why this is important: Shakespeare may have said that “The course of true love never did run smooth” and it would appear that he was right! Love and relationships have so many variables that it would appear that currently, data science is unable to predict outcomes - a humbling lesson that some things are still unknowable! [Click here to find out more!]( [Super Data Science podcast]( this week's [Super Data Science Podcast](, director of architecture at NVIDIA, Dr. Magnus Ekman, joins us to discuss the role of Machine Learning and Deep Learning in computer hardware. --------------------------------------------------------------- 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