Technology

Finding value from AI agents from day one

Source

MIT Technology Review

Published

Jul 17, 2025

TL;DR

AI Generated

Agentic AI, capable of autonomous decision-making, is gaining traction in corporate settings for tasks like procurement and cybersecurity. Despite excitement, experts caution against rushing deployment to avoid complexity and cost. Prioritizing interoperability early on can enhance the value of AI agents, with multi-agent systems expected to play a crucial role in the future. Organizations are advised to focus on incremental use cases and simplicity to maximize the benefits of agentic AI. Third-party developers are already offering multi-agent capabilities, but larger organizations may need to develop their own API architecture for long-term success.

Read Full Article

OpenAI is throwing everything into building a fully automated researcher

OpenAI is shifting its focus to building an AI researcher, aiming to create a fully automated system capable of tackling complex problems independently. The company plans to develop an autonomous AI research intern by September, leading to a multi-agent research system by 2028. OpenAI's chief scientist, Jakub Pachocki, believes in the potential of AI models to work autonomously for extended periods, with the goal of applying AI tools to real-world problem-solving. However, concerns about the risks and ethical implications of autonomous AI systems remain, prompting discussions on oversight and control mechanisms.

MIT Technology Review•

2 months ago

Agentic AI and the Future of Engineering

Agentic AI is highlighted in a Synopsys Converge keynote as a practical response to the growing complexity of engineering, particularly in the era of physical AI. This form of AI is seen as a complement to human engineers, enhancing their capabilities and allowing organizations to tackle more ambitious projects with limited resources. The keynote introduces a framework of autonomy levels from L1 to L5, showcasing a staged engineering roadmap where agentic AI plays a progressively more significant role in tasks and decision-making. By emphasizing collaboration between human engineers and AI systems, the keynote suggests a shift towards more strategic and creative engineering work. The model presented allows for customization and integration of customer-specific agents, data, and infrastructure, indicating a future of open, modular, and multi-model agentic AI systems in engineering.

SemiWiki•

2 months ago

Why iRobot’s founder won’t go within 10 feet of today’s walking robots

Rodney Brooks, a robotics pioneer and co-founder of iRobot, warns against standing within 9 feet of full-sized walking robots due to safety concerns. In his essay, he explains that current bipedal humanoids are unsafe for humans to be near when walking, as they generate significant kinetic energy that can cause harm if they fall or strike someone. Brooks believes that until safer versions of walking robots are developed, humanoid robots will not be certified for deployment in areas with people. He criticizes the billions being invested in humanoid development, calling it an expensive fantasy.

Ars Technica•

7 months ago

Undersea cable attacks drive sea drone development — Stark's Vanta unmanned vessels could be an affordable solution to protecting vital infrastructure

Stark has unveiled its new Vanta-4 and Vanta-6 unmanned surface vessels (USVs), designed to protect undersea cables from attacks. These sea drones are equipped with advanced sensors and have a range of up to 900 nautical miles. They can operate independently or in coordination with traditional navy vessels and aerial drones. Stark aims to leverage these drones for mass deployment to enhance NATO sea patrols. The company has received significant funding and support for its innovative solutions in maritime security.

Tom's Hardware•