Weekly roundup: The top AI stories you might have missed this week: - Microsoft said it will phase out facial recognition technology that's designed to identify a person's age, gender, emotional state, and other characteristics. The facial analysis capabilities are no longer available to new customers and will end for existing customers by June 30, 2023, unless for specially approved projects.
- The GitHub Copilot AI-powered code suggestion tool was made generally available to individual developers. The programming assistant costs $10 per month or $100 a year and is initially free for students.
- Similarly, Amazon launched CodeWhisperer, an AI-powered service that automatically generates software code. The ML-powered coding companion, which can generate more than 10 lines of code at a time, is now in preview.
- Instagram is now testing an AI-enabled facial analysis tool offered by the company Yoti to verify the ages of younger users. U.K.-based Yoti says it uses a neural network trained specifically for facial age estimation.
- Meta AI released a 66-billion-parameter language model for unrestricted download, its largest open-sourced NLP model released in this manner to date. Unlike its larger counterpart OPT-175B, OPT-66B does not require access request. The models are part of Meta's family of language models that can generate text, answer questions, solve math problems, and perform other tasks.
| |
OpenAI researchers trained neural networks to play Minecraft using mostly unlabeled videos of humans playing the game. The model can hunt, swim, cook animals, and engage in “pillar jumping" — a significant feat considering Minecraft's more open-ended gameplay. More: - OpenAI tapped a mix of unlabeled Minecraft videos and a small number of videos labeled by contractors to train an inverse dynamics model (IDM).
- The model is a mix of different neural nets that learn which actions go with which frames.
- In some cases, the IDM could perform on par with humans. It's also capable of more complex tasks like crafting "diamond tools" in the game, which typically takes adept players at least 20 minutes or about 24,000 actions.
- OpenAI researchers described their work in a paper and a blog post. They referred to their method, a type of semi-supervised imitation learning, as “Video PreTraining” or VPT.
- Alphabet's DeepMind, another AI lab, has trained its AlphaZero system to play Go, chess, and Shogi, and its subsequent MuZero program to master those games as well as Atari.
GIZMODO 
| |
 A message from BAMBOOHR Optimizing Compensation for Shared Trust It often takes a major disruption like the events of 2020 to highlight areas where the employer employee balance is more precarious than most people suspected. In this case, the disruption led to a free fall in retention we now call The Great Resignation—in November 2021 alone, more than 4.5 million U.S employees left their jobs. Understanding how these factors affect your employees’ productivity and your organization’s retention rate is essential for nurturing the effective partnerships you’ve built, not to mention growing into a better future together. Let’s explore how trust is built between employer and employee and what it needs to grow stronger in spite of today’s challenges. Download the Guide
| |
The Pentagon released a paper outlining how it plans to responsibly use AI. In the document, Deputy Secretary of Defense Kathleen Hicks said the U.S. needs to boost its AI investment to respond to China and other adversaries that are pumping more into AI technologies that “threaten global security, peace, and stability.” More: - Hicks said the DoD needs to embrace AI tech to "keep pace with these evolving threats."
- To do so responsibly, the 47-page document lists concrete ways to promote AI based on the Pentagon's six responsible AI "tenets."
- These include RAI governance, warfighter trust, AI product and acquisition lifecycle, requirements validation, responsible AI ecosystem, and AI workforce.
- The Pentagon said the new guidelines are meant to earn the trust of the public and service members.
Zoom-out: - In February 2020, then-U.S. Defense Secretary Mark Esper laid out the military's five ethical principles for using AI.
- The Pentagon said it wants its AI tech and efforts to be responsible, traceable, equitable, reliable, and governable.
- Last November, the Pentagon issued "responsible AI" guidelines that companies will have to follow when they develop and deploy AI technologies for the military. According to MIT Tech Review, the guidelines are designed to promote trust after Project Maven, which has focused on developing computer vision to analyze military surveillance footage.
GIZMODO
| |
AI investments in India are poised to grow at a compound annual growth rate (CAGR) of 30.8% and surpass $880M by 2023, according to a report issued by Nasscom, the National Association of Software and Services Companies. The AI Adoption Index forecasts that AI and data utilization could add $500B to India's GDP by 2025. More: - India's share of global AI investment is currently 1.5%. Nasscom predicts that will rise to 2.5% in 2023.
- Nasscom predicts that four sectors — healthcare, BFSI, consumer packaged goods and retail, and industrials/automotive — will contribute a combined 60% to the $500B opportunity.
- While enterprises in India have a long way to go to reach AI adoption maturity, 65% of organizations surveyed said they have defined AI strategy at a functional or an enterprise level, according to Nasscom.
- The report, which polled 350 companies, was published in partnership with Microsoft, Ernst & Young (EY), Microsoft, EXL, and Capgemini.
BUSINESS STANDARD
| |
 A message from MONDAY.COM Success and efficiency go hand in hand. And a big part of efficiency comes down to great teamwork. monday.com is the customizable work platform that empowers your team members to collaborate seamlessly and reach new heights, by simplifying and effectivizing workflows. Here’s how: - clever automations eliminate tedious tasks, such as reminders and status updates.
- ready-made templates let you plan, track, and manage everything from simple to complex projects - in one place.
- communication and collaboration in real-time on visually intuitive boards enables everyone to spend less time in meetings, and get more done.
- integrations connect your favorite tools, so you can continue using them all within the platform.
There are even more ways monday.com can help you and your team do your best work. Let us show you how. Show me how 
| |
AI contract-screening startup LegalForce raised a $101M Series D round led by SoftBank's Vision Fund. Japan-based LegalForce provides document analysis and contract management using ML and natural language processing. Its AI-based screenings can review contracts for potential legal issues, omissions, and loopholes. More: - The company says users can upload contracts to its "LegalForce Cabinet." Its AI system automatically scans through information like title, contract party name, and deadline, and creates a searchable ledger.
- It counts Nomura Holdings Inc. and All Nippon Airways Co. as among its 2,000 clients.
- Sequoia Capital China, Goldman Sachs Group Inc., Mitsubishi UFJ Capital, and Mizuho Capital also took part in the round, which brings LegalForce's fundraising total to $130M.
BLOOMBERG
| |
Fujifilm Irvine Scientific, a biotech firm, licensed an AI technology that evaluates human embryos in an effort to boost in-vitro fertilization success rates. Santa Ana, California-based Fujifilm picked up the technology from its developers, Boston-based Brigham and Women's Hospital and Massachusetts General Hospital. More: - The software platform taps AI and imaging to analyze embryos and identify the best candidates for IVF, improving success rates which are currently ~10% or less for women in their 40s.
- Fujifilm’s Irvine Scientific unit, which specializes in assisted reproductive technologies, intends to offer the technology to other U.S. fertility clinics through non-exclusive licenses.
- Terms of the deal weren't disclosed.
- Similarly, the fertility startup Alife Health recently raised a $22M Series A for its AI tools, which it says can lower the cost and improve the likelihood of successful in-vitro fertilization.
BECKER'S HOSPITAL REVIEW
| |
QUICK HITS - During uncertain times, customer-centric businesses stay afloat. Find out why in this deep dive.*
- GM-owned Cruise said it has started charging the public for robotaxi rides with no backup safety drivers in San Francisco.
- Three retailers in Australia — Bunnings, Kmart and The Good Guys — are using facial recognition technology in a bid to prevent theft, according to a Choice investigation. The consumer group found that the companies are "capturing highly personal data from customers."
- The National Highway Traffic Safety Association granted approval for Swedish freight company Einride to operate its self-driving electric pod trucks on U.S. public roads.
- Concerned about your SME’s security? Discover what a Zero Trust strategy entails – and how to use it to your advantage.*
*This is sponsored content.
| |
Upcoming Events: - June 30 - HR Strategies to Retain Remote Employees (Register Here)
- July 20 - The Rise of Metaverse Gaming (Register Here)
- July 27 - Need To Know NoCoder - Yassine Tahi (Kinetix) (Register Here)
- August 17 - Build Your Diversified Portfolio: How To Invest in Bear Markets (Register Here)
- September 14 - Are Electric Cars Worth It? (Register Here)
| |
 |  | Beth is a writer and analyst at Inside.com covering artificial intelligence and daily tech news. Since starting work at Inside, she has covered various topics including virtual reality, drones, and e-commerce. Prior to that she was a public policy and investigative reporter for The Arizona Republic, where she won a Pulitzer Prize nomination and First Amendment Award for reporting on the rising costs of pensions. Reach her at Beth.Duckett@inside.com. |   | | Editor | Gregory Bridgman is a writer and researcher with an academic background in politics and the philosophy of science and technology. He holds a BA from the University of Cape Town, an MS from University College London, and is currently completing a PhD at the University of Cambridge. He is interested in climate issues, technological changes, and the implications of the fourth industrial revolution. Please feel free to contact me at gregory@inside.com! | |
BambooHR: All-in-one software means oodles of features and a single source of data. | |
Boost your team’s efficiency and streamline workflows - with the help of monday.com. | |
Comments
Post a Comment