Author: The Horizon

Intel unveils new data center chip with focus on AI growth

Intel unveiled its latest generation of data center chips, the 5th Gen Intel Xeon processors, during its AI Everywhere event in New York.   The new chips offer AI functionality of their own and provide greater performance to customers deploying AI capabilities across cloud, network, and edge use cases.   The 5th Gen chips boast […]

Evaluating and Mitigating Discrimination in Language Model Decisions

Source: Anthropic Awhere they have not yet been deployed. Specifically, we use an LM to generate a wide array of potential prompts that decision-makers may input into an LM, spanning 70 diverse decision scenarios across society, and systematically vary the demographic information in each prompt. Applying this methodology reveals patterns of both positive and negative […]

Google’s most capable AI, Gemini, is now available for enterprise development

Google has announced that its generative AI model, Gemini, is now available to enterprises for app development in three sizes: Ultra, Pro, and Nano. The Pro version is accessible via API with certain usage limitations for free, and it offers features such as a 32K context window for text generation. Google has also introduced additional […]

Midjourney Alpha is here with AI image generations on the web

Midjourney, a widely used AI image-generating service, has launched the alpha version of its website, allowing users to create images directly on the site rather than through its Discord app. This release is currently limited to users who have generated over 10,000 images, with plans to expand accessibility in the near future. The web interface […]

Humana also using AI tool with 90% error rate to deny care, lawsuit claims

Humana, a major health insurance provider, is facing a lawsuit for allegedly utilizing an AI model with a 90 percent error rate to deny care to elderly people under their Medicare Advantage plans. The lawsuit claims that the use of this AI model constitutes a fraudulent scheme that results in financial gain for the company […]

Google’s new AI tool ‘MusicFX’ composes music with just a few words

The article discusses Google’s launch of MusicFX, an experimental tool using AI to generate music. The tool originates from Google’s AI Test Kitchen, a platform aiming to provide early access to its AI technologies for public feedback. MusicFX, while revolutionary, has limitations to protect original artists’ voices and styles. It also imposes layers of protection […]

Shock as news channel to use AI anchors next year

While AI advancements continue to revolutionize industries, the unveiling of plans by LA-based Channel 1 to launch AI news anchors has sparked concern among journalists and the public. Channel 1 aims to use AI-generated digital humans to deliver news updates on free streaming TV platforms, potentially disrupting the traditional news industry. The use of AI-generated […]

Human brain-like supercomputer with 228 trillion links coming in 2024

Australian scientists at the International Centre for Neuromorphic Systems at Western Sydney University are developing a neuromorphic supercomputer, named DeepSouth, that aims to simulate the synaptic operations of a human brain at full scale. This supercomputer will be capable of 228 trillion synaptic operations per second, which is comparable to the estimated number of operations […]

Tesla Unveils New Humanoid Robot

Tesla has unveiled its second-generation Optimus humanoid robot, showcasing its improved dexterity and agility, and potential to replace human workers in the company’s factories. The robot, featured in a video shared by Tesla CEO Elon Musk, can perform complex tasks such as flexing its fingers and hands, walking faster, doing squats, and picking up objects […]

Scientists unveil first complete cellular map of adult mouse brain

The article discusses the creation of a high-resolution atlas that charts neural neighborhoods for over 5,300 cell types in the entire mouse brain. The atlas is the result of a groundbreaking collaboration, revealing the brain’s complexity and diversity by combining single-cell RNA sequencing with spatial transcriptomics. It uncovers the deep connection between a cell’s genetic […]

Microsoft Packs Power Into Its New Small AI Model Phi-2

Microsoft Research has released Phi-2, a small language model (SLM) with 2.7 billion parameters, showcasing remarkable language understanding and reasoning abilities despite its smaller size compared to large language models (LLMs) like GPT and Gemini. Its efficient performance in specific tasks such as math and coding positions it as a viable alternative to larger models. […]

US Gov in talks with Nvidia over AI chip exports to China

The US government, led by the Biden administration, is closely monitoring the sale of Nvidia’s accelerators to China, specifically focusing on restricting the shipment of the most advanced AI chips that could be used for military applications. Commerce Secretary Gina Raimondo has emphasized the importance of Nvidia complying with export restrictions and selling GPUs for […]

What is Google’s Terrifying Gemini AI Project Ellmann?

Google is reportedly working on a new AI project called Project Ellmann, which aims to create an ultra-intelligent AI assistant capable of capturing a “birds-eye view” of users’ lives. This AI assistant, known as Ellmann Chat, is designed to ingest personal photos, documents stored in a Google account, and other contextual information to provide highly […]

Google unveils MedLM, a family of healthcare-focused generative AI models

Google has announced MedLM, a family of generative AI models fine-tuned for the healthcare industry, aimed at aiding healthcare workers in completing their tasks. The AI models are available to Google Cloud customers in the U.S. and can be used for a variety of medical tasks, such as drafting patient notes and identifying biomarkers. The […]

Braintrust’s seed round: $5m to build infrastructure for AI products

Braintrust, a company specializing in AI infrastructure, has secured a $5.1m seed round led by Saam Motamedi from Greylock, bringing their total capital raised to $8.3m. Their product, Braintrust, has emerged as an essential tool for AI development, serving leading technology companies like Zapier, Coda, Airtable, and Instacart. The company has experienced rapid growth within […]

Fully remote workers are at the highest risk of being replaced by AI, a Stanford professor says

The advancement of artificial intelligence (AI) is posing significant challenges and opportunities for the workforce, particularly for remote and hybrid workers. Nicholas Bloom, an economics professor at Stanford University, warns that fully remote roles, especially those that are repetitive and don’t require in-person interaction, are at the highest risk of being replaced by AI.   […]

Microsoft Training AI to Do Paperwork to Build Nuclear Power Plants

Microsoft is harnessing the power of AI to streamline the paperwork involved in setting up new nuclear power plants. The company is training a large language model with nuclear regulatory and licensing documents, aiming to expedite the process which can normally take years and cost hundreds of millions of dollars. By leveraging AI in nuclear […]

Tenyx aims to fix LLMs’ catastrophic forgetting problem

In the world of large language models (LLMs), fine-tuning them using domain-specific data is crucial for generating relevant outputs. However, this process can lead to “catastrophic forgetting,” where the model loses critical capabilities and reasoning skills. To address this issue, Tenyx is introducing a fine-tuning method that helps businesses adapt LLMs to their unique requirements […]