Independent evaluation and red teaming are critical for identifying the risks posed by generative AI systems. However, the terms of service and enforcement strategies used by prominent AI companies to deter model misuse have disincentives on good faith safety evaluations. This causes some researchers to fear that conducting such research or releasing their findings will result in account suspensions or legal reprisal. Although some companies offer researcher access programs, they are an inadequate substitute for independent research access, as they have limited community representation, receive inadequate funding, and lack independence from corporate incentives. We propose that major AI developers commit to providing a legal and technical safe harbor, indemnifying public interest safety research and protecting it from the threat of account suspensions or legal reprisal. These proposals emerged from our collective experience conducting safety, privacy, and trustworthiness research on generative AI systems, where norms and incentives could be better aligned with public interests, without exacerbating model misuse. We believe these commitments are a necessary step towards more inclusive and unimpeded community efforts to tackle the risks of generative AI.

Imagine the Colossus of Rhodes, the Statue of Zeus and the other ancient Seven Wonders of the World standing as they did thousands of years ago when first built. Artificial intelligence has done just that by recreating each historic structure in modern society with bustling tourists snapping photos with smartphones. Only one of the original seven survives today, with the others lost over time due to war, crumbling civilizations and natural disasters. But using the imagine generator Midjourney, AI has brought them back from the dead, allowing the world to take another look. Ancient artwork depicting the Colossus of Rhodes shows the statue straddling the harbor entrance, but researchers have determined such a feat would be impossible.

Recent advancements in Large Vision-Language Models (VLMs) have shown great promise in natural image domains, allowing users to hold a dialogue about given visual content. However, such general-domain VLMs perform poorly for Remote Sensing (RS) scenarios, leading to inaccurate or fabricated information when presented with RS domain-specific queries. Such a behavior emerges due to the unique challenges introduced by RS imagery. For example, to handle high-resolution RS imagery with diverse scale changes across categories and many small objects, region-level reasoning is necessary alongside holistic scene interpretation. Furthermore, the lack of domain-specific multimodal instruction following data as well as strong backbone models for RS make it hard for the models to align their behavior with user queries. To address these limitations, we propose GeoChat - the first versatile remote sensing VLM that offers multitask conversational capabilities with high-resolution RS images. Specifically, GeoChat can not only answer image-level queries but also accepts region inputs to hold region-specific dialogue. Furthermore, it can visually ground objects in its responses by referring to their spatial coordinates. To address the lack of domain-specific datasets, we generate a novel RS multimodal instruction-following dataset by extending image-text pairs from existing diverse RS datasets. We establish a comprehensive benchmark for RS multitask conversations and compare with a number of baseline methods. GeoChat demonstrates robust zero-shot performance on various RS tasks, e.g., image and region captioning, visual question answering, scene classification, visually grounded conversations and referring detection. Our code is available at https://github.com/mbzuai-oryx/geochat.

The security of port areas involves monitoring at various levels. What kind of ships are coming in, are they perhaps guilty of illegal fishing, and what cargo do they carry? Security officers and harbor masters often can't carry out these control duties all by themselves, which is why ports around the world are increasingly making use of smart surveillance systems to monitor maritime territory. TU/e researcher Amir Ghahremani developed new algorithms as well as a learning system to improve vessel identification. He will obtain his Ph.D. degree at the department of Electrical Engineering on Friday June 24.

The learning ability of artificial neural networks ("ANNs") falls under the scientific area of machine learning. Machine learning is a generic term for the artificial generation of knowledge from experience. More specific, an ANN learns from historical examples and can generalize these after the learning phase by learning the patterns contained in the examples. In machine learning, there are three learning paradigms. These include supervised and unsupervised learning as well as reinforced learning.¹

The installation process of offshore wind turbines requires the use of expensive jack-up vessels. These vessels regularly report their position via the Automatic Identification System (AIS). This paper introduces a novel approach of applying machine learning to AIS data from jack-up vessels. We apply the new method to 13 offshore wind farms in Danish, German and British waters. For each of the wind farms we identify individual turbine locations, individual installation times, time in transit and time in harbor for the respective vessel. This is done in an automated way exclusively using AIS data with no prior knowledge of turbine locations, thus enabling a detailed description of the entire installation process.

With tech giants announcing "AI-first" business models and investment in intelligent technology skyrocketing, it may come as a surprise to learn that only 20% of surveyed enterprises are actually implementing AI this year. While there's a prevailing sentiment among business leaders that they have missed the AI boat, the reality is that the boat is still in the harbor, and there's still plenty of room aboard. Early adopters of any technology do bystanders a great service by working out the kinks. Now that products have been tested and consumers introduced, contemporary businesses braving the AI waters will be known historically as the first generation to wield this promising technology to revolutionize industries. Certain companies are expecting returns on their AI investments to reach 30% in the coming years.

Autonomous vehicles have infiltrated much of the military, from airborne surveillance to all manner of ground-based operations. But the Navy remains a mostly human-controlled operation--with the demand for robotic tech focused on conflicts in Iraq and Afghanistan, it simply hasn't trickled down to aquatic operations yet. But the Office of Naval Research thinks autonomous boats can have a major impact on the military's ocean-going efficiency and effectiveness. In a demonstration conducted this fall in the lower Chesapeake Bay, a fleet of small, human-free boats collectively patrolled a harbor, detected intruders, and even chased them away from the area they were protecting. The Navy first demonstrated the swarm in 2014, when the vessels were tasked with protecting a single ship.