Latest news with #Guan
Yahoo
4 days ago
- Science
- Yahoo
Mathematicians Solve Multidimensional Shape-Slicing Dilemma
In 1986 Belgian mathematician Jean Bourgain posed a seemingly simple question that continued to puzzle researchers for decades. No matter how you deform a convex shape—consider shaping a ball of clay into a watermelon, a football or a long noodle—will you always be able to slice a cross section bigger than a certain size? A paper by Bo'az Klartag of the Weizmann Institute of Science in Rehovot, Israel, and Joseph Lehec of the University of Poitiers in France, posted to the preprint site has finally provided a definitive answer: yes. Bourgain's slicing problem asks whether every convex shape in n dimensions has a 'slice' such that the cross section is bigger than some fixed value. For three-dimensional objects, this is like asking whether an avocado of a given size, no matter the exact shape, can always be split into two halves with each side revealing at least some sizable slice. Bourgain, a titan of mathematics, is said to have spent more time on this problem than any other; although it may seem deceptively easy to resolve in the physical world's two or three dimensions, it quickly balloons in difficulty when we consider four or five. This added complexity makes determining anything in n-dimensional space seem impossible. 'If you believe in this so-called curse of dimensionality, you might just give up,' Klartag says. Fortunately, he adds, he and Lehec 'belong to a different school of thought.' The pair's breakthrough builds on recent progress by mathematician Qingyang Guan of the Chinese Academy of Sciences, who approached the problem with a technique based on physics rather than geometry. Specifically, Guan showed that modeling how heat diffuses out of a convex shape can reveal hidden geometric structures. Researchers could calculate filling any convex shape with warm gas and carefully observe the heat's dissipation according to physical laws. Guan's key insight—a precise limit on how rapidly the rate of dissipation changes during this heating process—proved to be just what Klartag and Lehec needed. 'Guan's bound tied together all the other key facts' known for the problem, says mathematician Beatrice-Helen Vritsiou of the University of Alberta. [Sign up for Today in Science, a free daily newsletter] The result let Klartag and Lehec resolve the problem in only a few days. Klartag notes that 'it was lucky because we knew [Guan's result] was exactly one of the things we needed' to connect several seemingly disparate approaches to the puzzle. With this final piece in place, the geometry of convex bodies in high dimensions is now a little less mysterious—although, as always in mathematics, each new slice reveals more questions to explore.
Scientific American
4 days ago
- Science
- Scientific American
Mathematicians Solve Multidimensional Fruit-Slicing Dilemma
In 1986 Belgian mathematician Jean Bourgain posed a seemingly simple question that continued to puzzle researchers for decades. No matter how you deform a convex shape—consider shaping a ball of clay into a watermelon, a football or a long noodle—will you always be able to slice a cross section bigger than a certain size? A paper by Bo'az Klartag of the Weizmann Institute of Science in Rehovot, Israel, and Joseph Lehec of the University of Poitiers in France, posted to the preprint site has finally provided a definitive answer: yes. Bourgain's slicing problem asks whether every convex shape in n dimensions has a 'slice' such that the cross section is bigger than some fixed value. For three-dimensional objects, this is like asking whether an avocado of a given size, no matter the exact shape, can always be split into two halves with each side revealing at least some sizable slice. Bourgain, a titan of mathematics, is said to have spent more time on this problem than any other; although it may seem deceptively easy to resolve in the physical world's two or three dimensions, it quickly balloons in difficulty when we consider four or five. This added complexity makes determining anything in n -dimensional space seem impossible. 'If you believe in this so-called curse of dimensionality, you might just give up,' Klartag says. Fortunately, he adds, he and Lehec 'belong to a different school of thought.' The pair's breakthrough builds on recent progress by mathematician Qingyang Guan of the Chinese Academy of Sciences, who approached the problem with a technique based on physics rather than geometry. Specifically, Guan showed that modeling how heat diffuses out of a convex shape can reveal hidden geometric structures. Researchers could calculate filling any convex shape with warm gas and carefully observe the heat's dissipation according to physical laws. Guan's key insight—a precise limit on how rapidly the rate of dissipation changes during this heating process—proved to be just what Klartag and Lehec needed. 'Guan's bound tied together all the other key facts' known for the problem, says mathematician Beatrice-Helen Vritsiou of the University of Alberta. On supporting science journalism If you're enjoying this article, consider supporting our award-winning journalism by subscribing. By purchasing a subscription you are helping to ensure the future of impactful stories about the discoveries and ideas shaping our world today. The result let Klartag and Lehec resolve the problem in only a few days. Klartag notes that 'it was lucky because we knew [Guan's result] was exactly one of the things we needed' to connect several seemingly disparate approaches to the puzzle. With this final piece in place, the geometry of convex bodies in high dimensions is now a little less mysterious—although, as always in mathematics, each new slice reveals more questions to explore.
Mint
03-05-2025
- Business
- Mint
AI agents are learning how to collaborate. Companies need to work with them
Developers are creating protocols to harness AI-powered agents into teams that handle everything from customer service and coding to supply chain, logistics, finance, marketing and business strategy. Companies should start planning for the next stage of artificial intelligence: the orchestration of multiple agents across their businesses. Most companies are still figuring out how to deploy even one AI-powered agent that can perform a task autonomously or in coordination with humans. But developers are creating protocols to harness these agents into teams that handle everything from customer service and coding to supply chain, logistics, finance, marketing and business strategy. Given the pace of innovation and the time it takes for organizations to adapt, companies will do themselves a favor by getting ready now for multiagent systems increasingly available later this year. Accenture's chief AI officer, Lan Guan, says only 10% to 15% of her clients currently use multiagent systems, but she expects that percentage to exceed 30% within 18 to 24 months. The professional services company has created a 15-agent system used for marketing, for example, comprising three 'super agents" that are responsible for coordinating 12 agents trained for specific tasks. It can plan a marketing campaign around a topic such as '2025 trends," conducting research, identifying similar past campaigns and addressing questions like a human would, according to Guan. All told, Accenture has more than 50 multiagent systems today for a range of industries and markets, and expects that number to hit more than 100 by the end of the year. The firm said customers such as carmaker BMW, consumer-brands company Unilever and sports giant ESPN are currently adopting these systems. Accenture last month introduced Trusted Agent Huddle, which it said allows agent-to-agent interoperability with partners such as technology companies Amazon Web Services, Google Cloud, Meta, Microsoft, Nvidia, Oracle, Salesforce, SAP and ServiceNow. Multiagent capabilities are about to become more widely available. Salesforce and Google announced at the Google Next conference in April that they were working on a protocol called A2A, or Agent-to-Agent. The protocol, which allows agents within Salesforce's Agentforce ecosystem to interact with each other as well as external agents, focuses on areas such as authentication, identification and message passing, according to Gary Lerhaupt, vice president of product architecture for Agentforce. Work is under way with partners to develop prototype multiagents using A2A, he said. Keyway, a commercial real-estate tech startup based in New York, provides a glimpse into how the concept works in practice, according to co-founder and Chief Executive Matias Recchia. It offers asset managers and property managers a multiagent platform that uses coordinated interactions to address questions such as how to price a rental property or target amenities and incentives. The company has raised $45 million from investors including Canvas Ventures, Camber Creek and Thomvest. While Keyway's agents are specialized and can trigger each other through structured workflows, the company said, they still operate in a controlled, predefined sequence that requires human oversight such as setting prompts, reviewing outputs and supervising decisions. A true multiagent system, Recchia said, involves agents that dynamically reason, negotiate or collaborate in real time without requiring human-defined workflows, explicit prompts or manual coordination. In other words, the agents take initiative, adapt to new information and interact fluidly with other agents without waiting for human instruction. Companies can start to prepare for multiagents systems simply by creating standard, stand-alone agents. Once the proper protocols are ready, companies can orchestrate these agents into tackling complex, collaborative systems. Principle Financial Group has embedded individual AI agents across domains including software engineering co-pilots , claims summarization and post-call analytics, according to Chief Information Officer Kathy Kay. They largely operate within defined scopes, but the investment management and insurance company is actively building the technical foundation to support agent-to-agent collaboration, Kay said. That means developing data pipelines and governance models . Workflows will also have to evolve to accommodate real-time collaboration between humans and intelligent, adaptive AI systems, she said. Kay foresees strong potential for the use of multiagent systems in retirement services such as rollover optimization. In asset management, she expects multiagents to analyze unstructured market data, generate investment narratives and align findings across portfolios. Other uses include earnings call preparation, contact center intelligence and claims adjudication. 'These are not isolated functions," Kay said. 'They are systems of tasks that, when connected through intelligent agents, can drive faster insights and better outcomes across the enterprise." Write to Steven Rosenbush at
Yahoo
15-03-2025
- Business
- Yahoo
The Best Way To Make Money With the 5 Most Popular AI Tools
According to The Listening App, 60% of Americans use AI-powered tools at least once weekly. The research also found the most common uses of AI apps are writing and editing, online research, summarizing text, brainstorming and generative art. You can even make money with these tools to boost your income. Learn More: Find Out: The Listening App listed the five most commonly used AI Tools and the percent of respondents who use them: ChatGPT (77.97%) Google Translate (44.89%) Gemini (33.23%) Canva (28.59%) Grammarly (25.33%) Here's a look at how you can make money with these popular AI-powered tools and what you need to know before getting started. 'You can make money with ChatGPT by making and selling prompt templates that are specific to certain industries or use cases,' said Shuai Guan, an AI expert and co-founder of Thunderbit. 'People can use these pre-written prompts to quickly and easily start writing high-quality content for marketing, customer service or even creative writing.' This revolutionary AI-powered tool has been a game changer since it hit the public market in late 2022, with many users finding different ways to make money from it. The best part of using ChatGPT is that you don't need a background in coding to use this tool, as you have your own personal assistant at your disposal. While creating prompts can be easy, you truly have to know the industry if you want to make sales. Another downside to trying to monetize ChatGPT is that there may not be a high demand for your services due to the ease of use. Check Out: Guan said you can earn anywhere from $50 to $500 per template pack sold. You can explore PromptBase to see what similar prompts are going for currently. You can also use ChatGPT to help you out with research as you try to make money with the other tools listed in this article. 'You can use Google Translate to tap into international markets with your product/service,' said Fergal Glynn, the chief marketing officer and AI security advocate for Mindgard. 'Apart from that, translate content or promote affiliate offers across different languages.' Chances are that you've used this app on your travels to get by and speak with locals. It turns out that you can also make money from it by translating your content into a different language. You'll want to review your analytics to see where your audience is coming from to determine if there are any gaps. The only downside is that translations may not always factor in local customs and your content may not flow as well as it does in your native language. You can diversify your affiliate income stream if you use translation services to convert your content into other languages. The average salary for affiliate marketing is $38 per hour, according to ZipRecruiter, but the earnings will depend on how well your content performs and the language that you choose to translate it to. 'Gemini can be leveraged to provide valuable competitive analysis services to businesses,' Guan said. 'By using Gemini's capabilities to gather and analyze data from various sources, individuals can offer insights into market trends, competitor strategies and opportunities for growth.' While there's money to be made using Gemini, it's important to note that offering AI-driven analysis will require a strong understanding of the specific industry and market research. This may not be the ideal AI-powered tool for beginners looking to make money. Guan noted that professionals with expertise in data analysis and industry knowledge could charge $75 to $150 per hour for their AI-driven competitive analysis services. 'Master using Canva to create logos, build websites and even design and sell templates,' Glynn said. 'Numerous independent contractors are earning passive income by selling their Canva templates on platforms like Etsy or through their own online stores.' Canva's design tools are easy to use, and anyone can make graphics from scratch. You don't need any specific design skills to get started, and you can create compelling images in minutes. The challenging part is figuring out what kind of products to offer, as you can make graphics, create social media posts or even sell digital templates. The notable negative of this tool is that it's becoming more popular, so demand for your graphics may not be as high if you're trying to build out a side hustle. These will vary based on the type of products that you create and how advanced your graphic design skills get. Guan said you could make anywhere from $25 to $75 per project, depending on how complex the projects are and how good you are at designing. If you use Canva to become a social media manager, for example, the average salary for this position is $64,845 annually, according to ZipRecruiter. The most logical way to earn money from Grammarly is to offer freelance writing services using its editing capabilities. You can also use Grammarly to provide proofreading and editing services, especially for academic or professional documents. The one setback of this side hustle is that it may not be ideal for everyone. Guan explained, 'If you want to use Grammarly to do academic editing and proofreading, you may need to know a bit about language, writing and editing.' Grammarly can help, but you need to know a lot about grammar and spelling and have at least some industry-specific knowledge. As a freelance writer, you can make $23 per hour on average, according to ZipRecruiter. The amount that you earn will depend on the outlets you write for. Many of these tools could be used together to deliver an exceptional finished product with the side hustle venture you choose to pursue. More From GOBankingRates10 Most Expensive Meals in the World10 Cars That Outlast the Average Vehicle This article originally appeared on The Best Way To Make Money With the 5 Most Popular AI Tools Sign in to access your portfolio
Yahoo
07-03-2025
- Business
- Yahoo
Chinese Lawmaker Floats More Secrecy to Help Evade US Chip Curbs
(Bloomberg) -- A Chinese lawmaker affiliated with a top local chip firm asked Beijing to let state-backed firms keep the identity of their foreign suppliers a secret, at a time Washington is increasingly blocking the country's access to Western hardware and semiconductors. Trump Administration Plans to Eliminate Dozens of Housing Offices NJ College to Merge With State School After Financial Stress Republican Mayor Braces for Tariffs: 'We Didn't Budget for This' How Upzoning in Cambridge Broke the YIMBY Mold NYC's Finances Are Sinking With Gauge Falling to 11-Year Low US-blacklisted, government-backed firms should be allowed to keep their suppliers anonymous when buying from overseas companies, National People's Congress delegate Guan Wenhui proposed this week. Instead of a public tender process as required by law, such corporations would then conduct purchases privately, she said. Publicizing the foreign sellers would only draw scrutiny and potentially force them to sever any remaining ties with Chinese firms, said Guan, who works for a unit of Naura Technology Group Co. — a chip gear maker that's been on a US export blacklist since late 2024. Guan's proposal shone a spotlight on the difficulties that Chinese tech buyers have grappled with since the US began a campaign to block the flow of technology to its geopolitical rival. US companies now cannot sell to blacklisted firms like Naura without first obtaining Washington's approval, forcing many foreign suppliers to cut ties with Chinese clients. In another instance, Nvidia Corp. is prohibited from supplying its most advanced chips to the country, stymieing China's AI development. In practice, many Chinese firms do find workarounds, for instance through third parties in neutral countries. 'So I would like to suggest these chip firms can be allowed to proceed with their procurement in a nonpublic manner to avoid disclosing their partners and ensure their access to advanced global tech,' Guan said during a breakout meeting on the sidelines of the annual NPC plenary session on Thursday. She also amplified an existing drive by Beijing to achieve self-sufficiency in critical technologies. While there's a process for turning proposals from NPC delegates into law, many such submissions never become concrete policies. Nonetheless, Guan's comments highlight a growing trend: Chinese data is becoming increasingly hidden from the world, making it harder and more costly to understand what's really happening in the country. Guan is a quality control engineer from Beijing 718 Yousheng Electronics Co., a firm that the US House Select Committee on China has identified as potentially having ties to the Chinese military. Yousheng is controlled by a unit of Naura, according to Chinese corporate data base Tianyancha. China's technology development has faced significant challenges after a yearslong US-led campaign to limit access to cutting-edge semiconductors, which power everything from artificial intelligence to smart vehicles. Beijing has said it will harness the entire nation's resources to achieve breakthroughs. The emergence of DeepSeek this year rallied China and fueled hopes that the country's AI progress isn't necessarily hampered by a lack of the most sophisticated chips. Still, many including Nvidia boss Jensen Huang argue that semiconductor firepower is key to winning the AI race. Snack Makers Are Removing Fake Colors From Processed Foods An All-American Finance Empire Drew Billions—and a Regulator's Attention The Mysterious Billionaire Behind the World's Most Popular Vapes Rich People Are Firing a Cash Cannon at the US Economy—But at What Cost? Greenland Voters Weigh Their Election's Most Important Issue: Trump ©2025 Bloomberg L.P.
