Awesome, not awesome.
“Not all robots are good at math. Take ProJo, a program that researchers are testing to help students of all ages spot their math and science mistakes, embodied in a small, humanoid robot. Instead of standing in for an instructor, ProJo acts as a peer, inviting the students themselves to help it solve problems. “Let’s take turns,” it might say. “I’m not so good at this.”
ProJo can also help students work together and assess their growth and weaknesses, in both robot form and on a computer screen. It is one of a variety of teaching aids in development, boosted by artificial intelligence, that scientists and educators say could support tomorrow’s classrooms.” — Sarah Trent and Aydali Campa, Reporting Interns Learn More from The Wall Street Journal >
“Natural language processing continues to find its way into unexpected corners. This time, it’s phishing emails. In a small study, researchers found that they could use the deep learning language model GPT-3, along with other AI-as-a-service platforms, to significantly lower the barrier to entry for crafting spearphishing campaigns at a massive scale.” — Lily Hay Newman, Senior Writer Learn More from WIRED >
What we’re reading.
1/ National Olympic teams have been using machine learning to analyze athlete data in order to gain any edge possible. Learn More from InfoQ >
2/ Machine learning models may be able to accurately predict terrorist attacks in areas that are already affected by terrorism. Learn More from The Next Web >
3/ A new project from Verily, an independent Google subsidiary, uses artificial intelligence in order to diagnose and treat colon cancer. Colon cancer currently kills over 900,000 people worldwide each year. Learn More from The Times of Israel >
4/ Banks begin to push for further use of artificial intelligence and machine learning to assist in making loan decisions. Many fear this will lead to additional unlawful discrimination and unfair practices. Learn More from The Columbus Dispatch >
5/ YouTuber uses machine learning to deter package thieves after recently having package stolen. Learn More from PC Gamer >
6/ A man from Germany spent 12 months “reclaiming his face” from Clearview AI, the controversial facial recognition company. Learn More from Coda >
7/ NASA is using machine learning for multiple space projects including Mars rovers, space medicine, robotic astronauts, and more. Learn More from Analytics Insight >
🙋♂️ Questions from the Community
This week we hosted an AMA on Slack with Jeffrey Ding, creator of the ChinAI newsletter. Through translating articles and documents from government departments, think tanks, traditional media, and newer forms of “self-media,” etc., ChinAI provides a unique look into the intersection between a country that is changing the world and a technology that is doing the same.
Below are some of his insights into the world of AI.
Q: What’s a topic related to China/AI strategy that is super interesting to you, but not currently reported on well?
A: One topic I’m drawn to are AI applications that are not as visible or consumer-facing as the ones you often hear and read about (think: machine quality inspection and machine translation as opposed to facial recognition). I call it Unsexy AI. For example, one of my favorite ChinAI issues was this translation of a longform article about China’s efforts to improve manufacturing lines for knives, peelers, and other cutting tools. The piece covered how Chinese factories are attempting to employ machine vision to detect defects in the production line, so as to improve manufacturing efficiency. I think this aspect of China’s AI development is really important because it will drive productivity growth, which is central to so much of the Chinese government’s broader strategic goals (sustained economic growth, performance legitimacy and regime stability, military capabilities, etc.)
Q: How do you grapple with the sometimes depressing/disturbing news that comes with AI-related news?
A: One thing I remind myself is that a lot of the depressing/disturbing news is not unique to covering this topic. That is, AI augments and exacerbates existing bad things in society, but it also provides an opportunity for redress and new paradigms.
Q: In the US, ML practitioners and increasingly even the mainstream media have grappled with the ethics of how AI/automation is applied. What does the conversation look like in terms of ethics/responsibility in China?
A: One way I like to conceptualize the AI ethics discussion is a red light-yellow light-green light system. Red=off limits/censored topics (e.g. use of facial recognition to discriminate against ethnic minorities like the Uyghurs). Yellow=some room for critique of govt surveillance like Dongyan Lao’s complaints about Beijing metro facial recognition but still murky area. Green = very robust discussions about AI and privacy, AI safety, and even risks associated with superinteligence or strong AI.
Q: How do you think the AI landscape in China will progress going forward? (more/less regulations?)
A: I predict more regulations. See, for example, the data security law and CAC regulations on deepfakes.
Q: What are the biggest misunderstandings that the US media currently has about China’s AI aspirations?
A: I would say US media consistently overestimate China’s actual AI capabilities. I cover others in my year #1 review post: https://chinai.substack.com/p/chinai-48-year-1-of-chinai
Are you interested in reading the full transcript from Jeffrey’s AMA? — join the growing Journal community here! (head on over to the #creator-amas channel!)
AI is able to write phishing emails better than humans was originally published in Machine Learnings on Medium, where people are continuing the conversation by highlighting and responding to this story.