Tag: DeepMind

DeepMind taps the power of its AI to accelerate quantum computers

Google DeepMind and UK-based Quantinuum have collaborated to advance the development of quantum computers by tackling a crucial challenge: reducing the number of T gates, which are vital but resource-intensive components in quantum circuits. Through their joint effort, they introduced AlphaTensor-Quantum, an AI model that combines deep reinforcement learning with tensor decomposition to optimize T-count. […]

Google DeepMind jumps back into open source AI race with new model Gemma

Google DeepMind has introduced Gemma, a set of 2B and 7B open-source models created using the same technology as the recent Gemini models. The Gemma models are accompanied by pre-trained and instruction-tuned variants and will be available under a permissive commercial license along with a new Responsible Generative AI toolkit. Google is offering toolchains for […]

Google DeepMind forms a new org focused on AI safety

Google’s Gemini AI model, known for generating realistic content based on prompts, has faced criticism for being capable of producing deceptive information, including false political content and fabricated details. This has drawn ire from policymakers and raised concerns about the misuse of AI tools for disinformation purposes.   In response to this backlash, Google, despite […]

Google DeepMind proposes ‘self-discover’ framework for LLMs, improves GPT-4 performance

Google DeepMind and the University of Southern California have unveiled an innovative ‘self-discover’ prompting framework to enhance reasoning capabilities. This morning’s publication on arXiV and Hugging Face introduces a revolutionary approach that surpasses existing techniques, significantly boosting the performance of esteemed models like OpenAI’s GPT-4 and Google’s PaLM 2.   Dubbed ‘self-discover,’ this cutting-edge framework […]

DeepMind Co-Founder: AI Is Fundamentally a “Labor Replacing Tool”

The founder of Google’s DeepMind, Mustafa Suleyman, acknowledged during an interview at the World Economic Forum that AI is fundamentally a “labor replacing tool.” This stance has become increasingly evident through recent deployment and the associated job losses.

DeepMind reveals AlphaGeometry: An Olympiad-level AI system for geometry

Trieu Trinh and Thang Luong introduce AlphaGeometry, an AI system that has achieved a significant breakthrough in solving complex geometry problems. The system’s performance approaches that of a human Olympiad gold-medalist, indicating a major advancement in AI reasoning and mathematical problem-solving.   AlphaGeometry utilizes a neuro-symbolic system, combining a neural language model with a symbolic […]

AMIE: A research AI system for diagnostic medical reasoning and conversations

AMIE is an AI system designed to engage in diagnostic conversations similar to those between clinicians and patients. AMIE is trained using a self-play based simulated learning environment and aims to improve diagnostic accuracy and conversation quality across various medical conditions and patient contexts. The study demonstrates that AMIE achieved diagnostic conversation performance at least […]

Isomorphic inks deals with Eli Lilly and Novartis for drug discovery

Isomorphic Labs, a drug discovery-focused spin-out of Google AI R&D division DeepMind, has partnered with pharmaceutical giants Eli Lilly and Novartis to leverage AI in discovering new medications to treat diseases, with combined deals valued at around $3 billion.   Isomorphic draws on DeepMind’s AlphaFold 2 AI technology to predict the structure of proteins in […]

Google wrote a “Robot Constitution” to make sure its new AI droids won’t kill us

In a groundbreaking move, DeepMind’s robotics team has introduced three innovative solutions to enhance robots’ decision-making capabilities in various environments, emphasizing safety, adaptability, and efficiency. The newly unveiled advances, including AutoRT with a visual language model (VLM) and large language model (LLM), the “Robot Constitution,” and safety features such as force threshold limits and a […]

Images altered to trick machine vision can influence humans too

In a groundbreaking study, research has shown that slight alterations to digital images, designed to deceive computer vision systems, also influence human perception. This discovery may highlight a resemblance between human and machine vision while emphasizing the necessity for further research to comprehend the impact of adversarial images on both humans and AI systems.   […]

Images altered to trick machine vision can influence humans too

In a groundbreaking study, research has shown that slight alterations to digital images, designed to deceive computer vision systems, also influence human perception. This discovery may highlight a resemblance between human and machine vision while emphasizing the necessity for further research to comprehend the impact of adversarial images on both humans and AI systems.   […]

AI can copy human social learning skills in real time, DeepMind finds

Researchers at DeepMind conducted a groundbreaking study showing that AI agents can demonstrate social learning skills by imitating humans in novel contexts without using pre-collected human data. In a simulated environment called GoalCycle3D, the AI agents learned to navigate by mimicking expert agents and quickly applied their new skills to other virtual paths. This advancement […]

Google DeepMind’s Demis Hassabis Says Gemini Is a New Breed of AI

Google has launched a new AI model called Gemini, which is integrated into the Bard chatbot. This AI model, as per Demis Hassabis, Google’s AI executive, represents a significant leap forward in the field of AI, as it is capable of processing information in various forms such as text, audio, images, and video. This multimodal […]

ChatGPT can leak training data, violate privacy, says Google’s DeepMind

The article discusses how researchers at Google’s DeepMind unit have found a new way to break the alignment of OpenAI’s ChatGPT, a generative AI program. By asking ChatGPT to repeat a word endlessly, the researchers were able to extract training data and even personally identifiable information.  

ChatGPT revealed personal data and verbatim text to researchers

In a surprising turn of events, researchers have discovered just how easy it is to extract personal information and training data from ChatGPT. The researchers, from esteemed institutions like Google DeepMind and Carnegie Mellon University, managed to uncover this vulnerability in OpenAI’s language model. They promptly informed OpenAI about their findings, and the issue has […]

Google DeepMind AI reveals potential for thousands of new materials

Google DeepMind has utilized artificial intelligence (AI) to predict the structure of over 2 million new materials, potentially improving real-world technologies such as batteries, solar panels, and computer chips. The AI firm stated that nearly 400,000 of these hypothetical material designs could soon be produced in laboratory conditions. The research, published in the science journal […]

©2023 The Horizon