A Baltimore high school athletic director was arrested on Thursday after police said he had allegedly used AI to create a fake audio clip of the school’s principal. In January, we reported on the audio clip that purported to be a recording of Pikesville High principal Eric Eiswert making racist and antisemitic comments about staff and students. At the time, Eiswert denied the authenticity of the audio, and claims were made that the clip was an AI fake. The widely shared audio clip led to Eiswert losing his job temporarily while the authenticity of the audio was investigated. The backlash from ..read more

Welcome to our roundup of this week’s freshest AI news. This week AI took blurry videos and made them look awesome. Microsoft released a tiny model that punches above its weight. And an AI pilot engaged humans in a real-world F-16 dogfight. Let’s dig in. Looking sharp Adobe teased us with demos of VideoGigaGAN, a model that upscales blurry video to look 8x sharper. We’ve seen attempts at video upscaling before but this is the first to generate smooth and sharp output from low-res videos. This could give us HD versions of historical archive footage and even HD versions of old TV shows. There’s ..read more

Microsoft launched Phi-3 Mini, a tiny language model that is part of the company’s strategy to develop lightweight, function-specific AI models. The progression of language models has seen ever larger parameters, training datasets, and context windows. Scaling the size of these models delivered more powerful capabilities but at a cost. The traditional approach to training an LLM is to have it consume massive amounts of data which requires huge computing resources. Training an LLM like GPT-4, for example, is estimated to have taken around 3 months and to have cost over $21m. GPT-4 is a great so ..read more

University of Illinois Urbana-Champaign (UIUC) researchers found that AI agents powered by GPT-4 can autonomously exploit cybersecurity vulnerabilities. As AI models become more powerful, their dual-use nature offers the potential for good and bad in equal measure. LLMs like GPT-4 are increasingly being used to commit cybercrime, with Google forecasting that AI will play a big role in committing and preventing these attacks. The threat of AI-powered cybercrime has been elevated as LLMs move beyond simple prompt-response interactions and act as autonomous AI agents. In their paper, the research ..read more

Radovan Kavicky is an AI & Data Science Evangelist at AIslovakIA, the National platform for AI development in Slovakia, and President of the GapData Institute. The non-profit platform supported by the Slovak government aims to connect all players in the AI industry in Slovakia and use the full potential of Slovakia’s greatest minds to promote AI development in Slovakia. Radovan completed his studies in Finance, so started as an Economist, but when you speak with him it’s clear that his real passion lies in Data Science & AI, especially as it relates to his home country of Slovakia. In ..read more

The Defense Advanced Research Projects Agency (DARPA) revealed that it performed the first-ever in-air test of an AI-piloted fighter jet in combat simulation against a human-piloted F-16. The demonstrations of the AI pilot’s flying prowess took place at the Air Force Test Pilot School at Edwards Air Force Base, California in September 2023 but were only announced by DARPA recently. The exercise saw the AI system developed by DARPA’s Air Combat Evolution (ACE) program pilot a specially modified F-16 test aircraft known as the X-62A, or VISTA (Variable In-flight Simulator Test Aircraft). We prev ..read more

Cogniteam operates in the IoT, robotics, and AI industries, delivering platforms to accelerate robotic management, deployment, and control.  Based in Petach Tikva, Israel, and with a US office in Boston, Cogniteam’s flagship product, the Cogniteam Platform, is a cloud-based system that enables the development, management, and control of robots, IoT devices, and fleets. Cogniteam’s platfrom enables comprehensive remote management, with a customizable dashboard that allows users to monitor in real time live data from anywhere in the world – regardless of how large and complex their fleet is ..read more

In a tumultuous day for the markets, NVIDIA’s stock tumbled by 10%, marking its worst performance since March 2020 and erasing over $200 billion from its market valuation.  This downturn contributed to the S&P 500 dropping by 0.9%, as outlined by Bloomberg data.  The broader market also felt the impact, with the tech-heavy Nasdaq Composite dropping by 2.1% upon close. Other major tech companies, including Advanced Micro Devices (AMD), Micron Technology, and Meta, saw their shares fall by 5.4%, 4.6%, and 4.1%, respectively. Super Micro Computer, noted for its server equipment bene ..read more

Welcome to our weekly roundup of ethically sourced AI news. This week AI taught robots how to play soccer. Udio turned everyone into a “musician”. And AI is advancing so fast that we need new ways to measure it. AI curious? Curiosity may have killed the cat, but behind every scientific discovery was a curious mind that thought, “I wonder what would happen if…” What would happen if we managed to develop an AGI that was curious? Could we do that? Should we? Sam’s fascinating exploration of the paradox of curiosity in the age of AI will challenge your opinion either way. Google scores! I’m guessi ..read more

The Oversight Board is an independent body of experts established by Meta in 2020 to review the company’s most challenging content moderation decisions on Facebook, Instagram, and Threads.  It announced on Tuesday that it will review Meta’s handling of two cases involving AI-generated sexually explicit images depicting female public figures. The board has 20 members worldwide, including lawyers, human rights advocates, journalists, and academics. It operates independently from Meta, with its own staff and budget, and can make binding decisions on content that Meta must implement unless do ..read more