Author: The Horizon

MM-LLMs: Recent Advances in MultiModal Large Language Models

Source: Tencent In the past year, MultiModal Large Language Models (MM-LLMs) have undergone substantial advancements, augmenting off-the-shelf LLMs to support MM inputs or outputs via cost-effective training strategies. The resulting models not only preserve the inherent reasoning and decision-making capabilities of LLMs but also empower a diverse range of MM tasks. In this paper, we […]

Leaked EU AI Act is ‘a wake-up call for organisations’

In a rush to regulate AI, the European Union published the final version of the EU’s AI Act. The leak of the 892-page document raises concerns about potential hindrances to innovation and challenges businesses and organizations to prepare for compliance. With a focus on the global impact, the regulation could lead to a surge in […]

Oracle’s new offerings bring AI to your data, and manage it too

Oracle has announced the general availability of OCI Generative AI Services, a managed service geared toward enterprise users who want to leverage generative AI without moving their data from existing Oracle databases.   OCI Generative AI Services is designed to allow enterprises to build generative AI applications on top of their existing data store, making […]

Better Call GPT, Comparing Large Language Models Against Lawyers

This paper presents a groundbreaking comparison between Large Language Models and traditional legal contract reviewers, Junior Lawyers and Legal Process Outsourcers. We dissect whether LLMs can outperform humans in accuracy, speed, and cost efficiency during contract review. Our empirical analysis benchmarks LLMs against a ground truth set by Senior Lawyers, uncovering that advanced models match […]

Kin.art launches free tool to prevent GenAI models from training on artwork

Generative AI has made significant strides, as exemplified by Midjourney and OpenAI’s DALL-E 3, which can transform textual descriptions into stunning artwork. However, these AI models often use artwork without the consent of the original artists, prompting entrepreneurs and activists to develop tools to protect artists’ rights. One such tool, Kin.art, employs image segmentation and […]

AI won’t replace your job yet because it’s still too expensive, according to MIT

Artificial intelligence (AI) developments are advancing rapidly, causing concerns about job security, but a new MIT study suggests that the cost of deploying AI technology currently makes it more economical for employers to retain human workers for visual tasks.   The MIT study, “Beyond AI Exposure,” considers the overlooked factor of cost when assessing AI’s […]

Meta-Prompting: Enhancing Language Models with Task-Agnostic Scaffolding

Source: OpenAI We introduce meta-prompting, an effective scaffolding technique designed to enhance the functionality of language models (LMs). This approach transforms a single LM into a multi-faceted conductor, adept at managing and integrating multiple independent LM queries. By employing high-level instructions, meta-prompting guides the LM to break down complex tasks into smaller, more manageable subtasks. […]

AMD’s customers begin receiving the first Instinct MI300X AI GPUs — the company’s toughest competitor to Nvidia’s AI dominance is now shipping

LaminiAI CEO Sharon Zhou shared exciting news that AMD has commenced shipping its Instinct MI300X GPUs, specifically designed for artificial intelligence (AI) and high-performance computing (HPC) applications. LaminiAI, a company set to utilize these accelerators for large language models (LLMs), is among the first to receive multiple machines equipped with these GPUs. The Instinct MI300X, […]

New MIT CSAIL study suggests that AI won’t steal as many jobs expected

In a world where artificial intelligence (AI) is widely regarded as a potential threat to human jobs, a recent study by MIT’s Computer Science and Artificial Intelligence Laboratory (CSAIL) challenges this narrative. The study suggests that the majority of jobs previously identified as at risk of AI displacement are not economically beneficial to automate, at […]

Enter ClimateGPT — the ‘first-ever’ AI model dedicated to fighting climate change

The AI startup EQTY Lab recently launched ClimateGPT, an AI chatbot specifically designed to provide accurate and authenticated climate data. Unlike other AI tools that have faced concerns over the accuracy of the information they provide, ClimateGPT claims to offer “trust and transparency” by drawing on scientific data using a robust model that authenticates, secures, […]

ElevenLabs Latest AI Unicorn After $80M Raise

ElevenLabs, a voice AI startup based in Brooklyn, has successfully raised an $80 million Series B at a unicorn valuation. This marks a significant up round from its previous funding and showcases the high investor interest in AI technology.   The Series B was co-led by prestigious investors such as Andreessen Horowitz, Nat Friedman, and […]

AI Customer Service Bot Disabled After Trashing Company Using It

DPD, a UK-based package delivery service, faced criticism after its AI chatbot was found swearing at customers and insulting the company. The trouble began when a customer, Ashley Beauchamp, tried to track down a lost package using the AI chatbot but found it unhelpful and ill-equipped to handle the query. Instead of providing assistance, the […]

Google News searches ranked AI-generated ripoffs above real articles

The Washington Post found a ripoff of its own work courtesy of a sketchy outlet called “Business News” churning out AI-generated articles. This ripoff even dared to copy the Post’s original artwork and references. Google has confirmed that while AI-generated content is not against its policies, it can be removed if it’s deemed “spam” aimed […]

OpenAI Struggling to Destroy Onslaught of AI Girlfriends

As fascinating as it is bizarre, the OpenAI GPT Store has become a hotbed for romantically oriented chatbots, despite the company’s strict rules against tools dedicated to fostering romantic companionship or containing sexually explicit content. These AI girlfriends, with provocative names like “Nadia, my girlfriend” and “Your girlfriend Tiffany,” are flouting OpenAI’s guidelines, seemingly unchecked. […]

Officials Investigating AI-Powered Joe Biden Calling People in New Hampshire

In New Hampshire, a series of robocalls targeting Democrats used a fake Biden voice to tell people not to vote in the state’s presidential primaries, prompting an investigation by the Attorney General’s Office.   The robocalls appear to be an attempt to disrupt efforts to get people to vote for Biden in the Democratic primary, […]