While Multimodal Large Language Models (MLLMs) offer strong perception and reasoning capabilities for image-text input, Visual Question Answering (VQA) focusing on small image details still remains a challenge. Although visual cropping techniques seem promising, recent approaches have several limitations: the need for task-specific fine-tuning, low efficiency due to uninformed exhaustive search, or incompatibility with efficient attention implementations. We address these shortcomings by proposing a training-free visual cropping method, dubbed FOCUS, that leverages MLLM-internal representations to guide the search for the most relevant image region. This is accomplished in four steps: first, we identify the target object(s) in the VQA prompt; second, we compute an object relevance map using the key-value (KV) cache; third, we propose and rank relevant image regions based on the map; and finally, we perform the fine-grained VQA task using the topranked region.

As soon as Echo Park Eats opened on the corner of Sunset Boulevard and Douglas Street in the fall of 2023, Sandy Romero said her neighborhood became overrun with delivery drivers. "The first day that they opened business it was chaotic, unorganized and it's just such a nuisance now," she said. Echo Park Eats is a ghost kitchen, a meal preparation hub for app-based delivery orders. It rents its kitchens to 26 different food vendors. The facility is part of CloudKitchens, led by Travis Kalanick, co-founder of Uber Technologies, which has kitchen locations across the nation including 11 in Los Angeles County.

The term artificial intelligence was first used in a 1955 proposal for a study submitted by John McCarthy of Dartmouth College, Marvin Minsky of Harvard University, Nathaniel Rochester at IBM, and Claude Shannon at Bell Telephone Laboratories. This happened before I was born. I find it kind of nuts that AI was discussed long before we had the computing and storage power needed to make it work. As a decision support analyst fresh out of college, I built early AI systems that were too expensive to operate, so they were only used in specialized circumstances. It was a niche technology.

Something incredible is happening in AI right now, and it's not entirely to the good. Everybody is talking about systems like chatGPT (OpenAI), Dall-E 2, and Lensa that generate text and images that look remarkably human-like, with astonishingly little effort. These systems can be incredibly fun to play with. Take this example, generated by chatGPT by Henry Minsky (son of Marvin Minsky, one of AI's founders), who asked chatGPT to "Describe losing your sock in the dryer in the style of the declaration of independence": When in the course of household events, it becomes necessary for one to dissolve the bonds that have connected a sock to its mate, and to assume among the powers of the laundry room, the separate and equal station to which the laws of physics and of household maintenance entitle it, a decent respect to the opinions of socks requires that it should declare the causes which impel it to go missing. We hold these truths to be self-evident, that all socks are created equal, and are endowed by their manufacturer with certain unalienable rights….

While it might sound a bit reductive, the essential skill of real estate development can often be boiled down to one word: decisions. A project's developer might have more than 200 vendors to communicate with regarding pricing, timing, materials and a slew of other factors, juggling spreadsheets and details that could throw a project off its timeline and budget if this information isn't easy to source and reference when making project decisions. Keeping track of this multitude is the first step on the road to success for any developer. With the evolution of modern technology -- specifically, artificial intelligence (AI) -- developers are finding that the drudgery of tracking and inputting thousands of essential details is going the way of the rotary phone. The adoption of AI by the commercial real estate (CRE) industry was greatly accelerated by the uncertainty caused by the COVID-19 pandemic, as commercial real estate stakeholders embraced AI's ability to bring solutions into a clearer view.

A MACHINE with human-level intelligence could be built in the next 30 years and could represent a threat to life on Earth, some experts believe. AI researchers and technology executives like Elon Musk are openly concerned about human extinction caused by machines. The Law of Accelerating Returns is a concept popularized by futurist Ray Kurzweil that states the rate of technological improvement is on a very steep curve. As technology gets more advanced, society and industry are better equipped to improve technology faster and more drastically. "With more powerful computers and related technology, we have the tools and the knowledge to design yet more powerful computers, and to do so more quickly," Kurzweil wrote in his famous 2001 essay.

Though machine learning has the power to dramatically change healthcare, physicians should not fear being overrun by robots, according to the Harvard Business Review. Although machine learning can prevent patients from getting sick and diagnose a patient, the software can't provide care and treatment to patients. Additionally, machine learning lacks the human element in healthcare. While having data can be powerful, not all data is sufficient and necessary. Healthcare organizations must collect the right data and fully understand it.

In a suburb of Phoenix, Arizona, there's a fleet of 600 minivans shuttling people from place to place. Ordering one feels almost exactly like calling an Uber, except for one thing: the vans are driving themselves. Alphabet's Waymo has been testing self-driving vehicles in Arizona since 2017 and we got an inside look at what it's like. Watch the video to find out more.

The AI alliance aims to bring a new paradigm of data-driven project management to better manage construction projects of any kind all over the world. Major projects regularly underperform due to traditional systems and human limitations meaning outcomes are inaccurately predicted. The increasing complexity of modern-day large infrastructure design and construction means that major projects will continue to overrun unless a new approach is taken. Dan Phillips, global practice lead for project, programme and commercial management at global engineering, management and development consultancy Mott MacDonald, said: "There is a huge amount of money and time invested in the delivery of major projects, and despite this, they continue to underperform, deliver less than budgeted benefits and inevitably become the focus of political and media scrutiny. "The construction industry lags behind other industries in the use of data and adoption of AI technology will play a big part in helping us achieve better outcomes for project teams, clients and ultimately the end-users who feel the effect of project delays and cost overruns the most." Endeavour Programme, a technology firm headquartered in Brisbane, has developed Octant AI, a scalable cloud hosted case-based decision-making system. It gathers data and analyses the performance in construction projects to produce improved insight allowing earlier, more effective, decision making. The new alliance will use this technology. David Porter, Endeavour programme founder, commented: "Large projects have become so complex with so many interdependencies and stakeholders, that managing them with traditional methods is ever more challenging.