In an incredibly shocking turn of events, more generative AI features are coming to the Google Workspace suite of productivity apps by way of Gemini. Docs, Sheets, Chat, Meet and newcomer Vids are among the tools that are getting new AI capabilities that perhaps some folks will find useful. Google is taking a cue from audio overviews in NotebookLM to bring new Gemini-powered audio features to Docs. You'll soon have the option of listening to full audio versions of your documents or just the highlights in what Google calls a podcast-style overview. Those enrolled in the Workspace alpha testing program can start trying out these audio features in the coming weeks.

This work investigates the capabilities of large language models (LLMs) in detecting and understanding human emotions through text. Drawing upon emotion models from psychology, we adopt an interdisciplinary perspective that integrates computational and affective sciences insights. The main goal is to assess how accurately they can identify emotions expressed in textual interactions and compare different models on this specific task. This research contributes to broader efforts to enhance human-computer interaction, making artificial intelligence technologies more responsive and sensitive to users' emotional nuances. By employing a methodology that involves comparisons with a state-of-the-art model on the GoEmotions dataset, we aim to gauge LLMs' effectiveness as a system for emotional analysis, paving the way for potential applications in various fields that require a nuanced understanding of human language.

Imaging of facial affects may be used to measure psychophysiological attributes of children through their adulthood, especially for monitoring lifelong conditions like Autism Spectrum Disorder. Deep convolutional neural networks have shown promising results in classifying facial expressions of adults. However, classifier models trained with adult benchmark data are unsuitable for learning child expressions due to discrepancies in psychophysical development. Similarly, models trained with child data perform poorly in adult expression classification. We propose domain adaptation to concurrently align distributions of adult and child expressions in a shared latent space to ensure robust classification of either domain. Furthermore, age variations in facial images are studied in age-invariant face recognition yet remain unleveraged in adult-child expression classification. We take inspiration from multiple fields and propose deep adaptive FACial Expressions fusing BEtaMix SElected Landmark Features (FACE-BE-SELF) for adult-child facial expression classification. For the first time in the literature, a mixture of Beta distributions is used to decompose and select facial features based on correlations with expression, domain, and identity factors. We evaluate FACE-BE-SELF on two pairs of adult-child data sets. Our proposed FACE-BE-SELF approach outperforms adult-child transfer learning and other baseline domain adaptation methods in aligning latent representations of adult and child expressions.

Schmidt imagined a near future where China and the U.S. needed to cement a treaty around AI. "In the 50s and 60s, we eventually worked out a world where there was a'no surprise' rule about nuclear tests and eventually they were banned," Schmidt said. I'm very concerned that the U.S. view of China as corrupt or Communist or whatever, and the Chinese view of America as failing…will allow people to say'Oh my god, they're up to something,' and then begin some kind of conundrum. Begin some kind of thing where, because you're arming or getting ready, you then trigger the other side. We don't have anyone working on that and yet AI is that powerful."

Researchers have combined speech and facial recognition data to improve the emotion detection abilities of AIs. The ability to recognise emotions is a longstanding goal of AI researchers. Accurate recognition enables things such as detecting tiredness at the wheel, anger which could lead to a crime being committed, or perhaps even signs of sadness/depression at suicide hotspots. Nuances in how people speak and move their facial muscles to express moods have presented a challenge. Detailed in a paper (PDF) on Arxiv, researchers at the University of Science and Technology of China in Hefei have made some progress.

You know the type: Those friends who think they've found the best way to game Tinder, insist wearing a dress instead of jeans will catch someone's attention, or that the most fruitful way to find love is by going on the Bachelor. There is so much dating lore and folk wisdom out there, but we've never really had hard data behind it--at least until now. A group of researchers recently attempted to decode the world of heterosexual online dating and found not just that women go for older men and men for younger women, but that so many Americans are seeking a partner "out of their league." On average, researchers found, both heterosexual men and women go after people who are around 25 percent more "desirable" than themselves. But, you're probably wondering, how the hell are they defining what makes someone "desirable"?

Automation is bringing change to businesses large and small. We sat down with Liz Reynolds, executive director of MIT's Work of the Future task force, to see how her research brings her in touch with small companies in the US that are already feeling the impact of robotics and AI. This piece appeared in our daily newsletter Clocking In, which covers the future of work. How is automation affecting smaller companies? We've been out talking to small and medium-sized manufacturing firms, and what's exciting is that this automation technology is now accessible to them.

Artificial intelligence (AI) is getting a lot of attention from investors these days and with good reason. It's a technology with the potential to transform everything from transportation to cloud computing. Many companies are betting that AI will influence not just their products and services, but nearly every aspect of their business. Read on to discover nine things you may not have to know about this burgeoning technology. Artificial intelligence is often used as a catch-all to describe computer systems that can learn.

The AAAI News Column covers membership and conference statistics, reports on workshops, financial data, Council activities, and other ongoing activities pursued by the society Please send suggestions and feedback by electronic mail to AAAI-Office@Sumex-Aim Stanford.edu Spring Symposium Series In March AAAI sponsored the first Spring Symposium Series, intended to promote technical discussion of specialty topics in small-group meetings. As the following reports testify, the Series was a tremendous success Plans are underway for the next Series, to be held at Stanford University again next March. Suggestions should be sent to Hector Levesque or Claudia Mazzetti. The following are edited reports provided by the five symposium leaders.

When IBM's Deep Blue chess computer defeated the world champion Garry Kasparov in 1997, humanity let out a collective sigh, recognising the loss of an essential human territory to the onslaught of thinking machines. And not just against them: for the past two decades, Kasparov has been exploring an idea he calls'Advanced Chess', where humans collaborate with computer chess programs against other hybrid teams, sometimes called'Centaurs'. We rely on computational systems for our essential aesthetic vocabulary, learning what is good and beautiful through a prism of five-star rating systems and social-media endorsements, all closely watched over by algorithmic critics of loving grace. Many artists today explore the seams and rough edges of digital platforms, creating art out of the glitches and unintended juxtapositions that they can eke out of increasingly complicated creative systems.