Toggle light / dark theme

For their study published in the journal Nature Medicine, the group generated thousands of articles containing misinformation and inserted them into an AI training dataset and conducted general LLM queries to see how often the misinformation appeared.

Prior research and anecdotal evidence have shown that the answers given by LLMs such as ChatGPT are not always correct and, in fact, are sometimes wildly off-base. Prior research has also shown that misinformation planted intentionally on well-known internet sites can show up in generalized chatbot queries. In this new study, the research team wanted to know how easy or difficult it might be for malignant actors to poison LLM responses.

In today’s AI news, for those who were thinking Meta Platforms Inc. might back down from its heavy-spending ways in the wake of the DeepSeek news, think again. On the earnings call, Zuckerberg spoke of “the hundreds of billions of dollars” Meta will invest in AI infrastructure over the long term.

S ChatGPT and the newly ascending DeepSeek. “Qwen 2.5-Max outperforms… almost across the board GPT-4o, DeepSeek-V3 and Llama-3.1-405B,” Alibaba Meanwhile, amid the market turbulence and breathless headlines, Dario Amodei, co-founder of Anthropic and one of the pioneering researchers behind today’s large language models, published a detailed analysis that offers a more nuanced perspective on DeepSeek’s achievements.

And, Microsoft is bringing Chinese AI company DeepSeek’s R1 model to its Azure AI Foundry platform and GitHub. The R1 model is now part of a model catalog on Azure AI Foundry and GitHub — allowing Microsoft’s customers to integrate it into their AI applications.

In videos, Reid Hoffman is the co-founder of LinkedIn, a legendary Silicon Valley investor, and author of the new book Superagency: What Could Possibly Go Right with Our AI Future. Hoffman joins Big Technology Podcast to discuss his optimistic case for AI, the massive investments flooding into the field, and whether they can pay off.

And, in the new episode of Moonshots, Emad Mostaque, Salim Ismail, and Peter Diamandis discuss recent DeepSeek news, the China vs. USA AI race, and more. Emad is the founder of Intelligent Internet and former CEO of Stability AI. Salim is a serial entrepreneur and strategist well known for his expertise in exponential organizations.

Then, reverse the diffusion process, and unlock the secrets of AI-generated images. Isaac Ke explores how to harness the power of diffusion models to create stunning, high-quality images from text prompts. From text-to-image generation to image-to-image editing, learn how diffusion models are being applied in various fields.

We close out with, Nate B. Jones sharing the strategy behind DeepSeek strategically launching free AI products to disrupt OpenAI’s dominance, including an Operator clone, voice AI, and a likely Sora competitor. This aggressive move threatens OpenAI’s paid model and forces competitors to rethink pricing.

Electromagnetic absorbers are essential in energy, stealth, and communication technologies, yet current designs underperform. A research team has introduced ultra-thin absorbers nearing theoretical efficiency limits, promising transformative industrial applications.

Absorbing layers are essential to advancements in technologies like energy harvesting, stealth systems, and communication networks. These layers efficiently capture electromagnetic waves across wide frequency ranges, enabling the creation of sustainable, self-powered devices such as remote sensors and Internet of Things (IoT) systems. In stealth technology, absorbing layers reduce radar visibility, enhancing the performance of aircraft and naval systems. They also play a vital role in communication networks by minimizing stray signals and mitigating electromagnetic interference, making them indispensable in today’s interconnected world.

On the positive side, some human entrepreneurs could become very wealthy, possibly trillionaires if they could tap into these AI’s wealth somehow. Additionally, super rich AIs could be a solution to the United States’ growing debt crisis, and eliminate the need for whether countries like China can continue to buy our debt so we can indefinitely print dollars. In fact, can America launch its own AI agents to create enough crypto wealth to buy its debt?

Naturally, the risk is that these AIs might eventually try to buy other financial instruments, like existing bonds and stocks. But it’s unlikely they’d be able to do so, unless more of the U.S.’ economy went into crypto and became blockchain based. Additionally, AI bots aren’t allowed to have traditional bank accounts yet.

Whatever happens, clearly there is an urgent need for the U.S. government to address such potentialities. Given that these AIs could start to proliferate in the next few months, I suggest Congress and the Trump administration immediately convene a special task force to specifically tackle the possibility of an AI Monetary Hegemony.

The real danger is that even with regulation, programmers will still be able to release autonomous AIs into the wild—just as many illegal things already happen on the web despite the existence of laws. Programmers might release these types of AIs for kicks, while others try to profit from it—and some may even do so even as a form of terrorism to try to hamper the world economy. Whatever the reason, the creation of autonomous AIs will soon be a reality of life. And vigilance and foresight will be needed as these new AIs start to autonomously disrupt our financial future.

The world of quantum physics is experiencing a second revolution, which will drive an exponential leap in the progress of computing, the internet, telecommunications, cybersecurity and biomedicine.

Quantum technologies are attracting more and more students who want to learn about concepts from the subatomic world—such as quantum entanglement or —to explore the innovative potential of quantum science.

In fact, understanding the non-intuitive nature of quantum technology concepts and recognizing their relevance to technological progress is one of the challenges of 2025, declared the International Year of Quantum Science and Technology by UNESCO.

En un esfuerzo por fortalecer la ciberseguridad de los consumidores, la Comisión Federal de Comunicaciones (FCC, por sus siglas en inglés) ha lanzado el programa U.S. Cyber Trust Mark. Esta iniciativa voluntaria tiene como objetivo ayudar a los consumidores a identificar dispositivos del Internet de las Cosas (IoT) que cumplan con estrictos estándares de ciberseguridad, al tiempo que incentiva a los fabricantes a adoptar mejores prácticas. Sin embargo, se aplican ciertas restricciones, especialmente en lo que respecta a productos de algunas entidades extranjeras, incluidas empresas chinas.

Los dispositivos inteligentes, como cámaras de seguridad para el hogar, rastreadores de actividad física y electrodomésticos inteligentes, se han convertido en elementos esenciales de la vida moderna. Aunque ofrecen una comodidad sin precedentes, estos dispositivos también presentan riesgos significativos de ciberseguridad, haciéndolos vulnerables a ataques informáticos.

El programa Cyber Trust Mark busca mitigar estos riesgos proporcionando una etiqueta que indique que un producto cumple con estándares robustos de ciberseguridad. La presidenta de la FCC, Jessica Rosenworcel, destacó la importancia de la iniciativa, afirmando: “Este programa no solo ayuda a proteger a los consumidores, sino que también crea incentivos para que los fabricantes prioricen la ciberseguridad.”

A team of researchers from the University of Ottawa has developed innovative methods to enhance frequency conversion of terahertz (THz) waves in graphene-based structures, unlocking new potential for faster, more efficient technologies in wireless communication and signal processing.

THz waves, located in the far-infrared region of the electromagnetic spectrum, can be used to perform non-invasive imaging through for security and quality control applications. Additionally, these waves hold great promise for wireless communication.

Advances in THz , which can be used to change the frequency of electromagnetic waves, are essential for the development of high-speed and signal processing systems for 6G technologies and beyond.