True AI Transformation requires a pioneering spirit, precision, and the courage to traverse the unknown

Let's go

An Applied AI Studio by Manifold unlocking opportunities for growth & transformation.

Learn More about our labs-based approach to unlocking potential.

A curated collection shared by our team.

New paper proposes infinite sequence lengths without any fine tuning

This article introduces StreamingLLM, an efficient framework that allows Large Language Models (LLMs) to generalize to infinite sequence lengths in streaming applications without fine-tuning, addressing the challenges of extensive memory consumption and poor performance when the text length surpasses the cache size.
Play

Exploring potential of LLMs in Scientific Experimentation

This article explores how Large Language Models (LLMs) can enhance scientific experimentation by improving experimental design, implementation, and data analysis, thus increasing the likelihood of reporting accurate findings.
Play

Meta launches new AI assistant, glasses, and Quest 3

Mark Zuckerberg announced new AI products, including photo-realistic image bots, smart glasses, and an updated virtual reality headset, at Meta Connect, Meta Platforms' largest event since the pandemic; the new products aim to integrate virtual and real worlds, offering affordable and free AI integration.
Play

Amazon Invests $4 Billion in Anthropic to Advance Safe and Controllable AI

Amazon will invest up to $4 billion in AI startup Anthropic, gaining early access to its technology and reinforcing its presence in the AI industry.
Play

Just connect your data sources and get a ChatGPT-like chatbot for your data.

Chatbase, a chat analytics platform, now integrates with popular tools to provide seamless data analysis and improve chatbot performance.
Play

Voice and image support dropping soon for ChatGPT Plus + Enterprise accounts

OpenAI is introducing voice and image capabilities to ChatGPT, allowing users to have voice conversations and share images, aiming to provide a more intuitive experience; however, precautions are in place to mitigate risks such as impersonation and misuse, and users are encouraged to provide feedback to improve the system.
Play

Multi Agent Collaboration

ChatDev is an open-source project using multiple AI agents in a game-like UI to collaboratively build software. This method, supported by multiple studies improves AI abilities. There's little doubt this is just the start of what we'll see coming over the next year.
Play

DALL-E 3 is now in research preview

The Image generation model from OpenAI aims to allow refined image prompts, via direct integration with ChatGPT. DALL·E 3 will be accessible to ChatGPT Plus and Enterprise customers in October, and users retain full rights to the images they produce.
Play

Making over assignments, educators and students collaborate on AI ground rules

Teachers are struggling to adapt assignments to prevent students from using AI chatbots like ChatGPT to complete their homework, with some resorting to banning the use of such tools, while others focus on making assignments more interesting and authentic to deter cheating.
Play

A new study proposes 14 functional requirements to mitigate risk of AI harm to patients in Healthcare

The report highlights seven risks of AI in medicine and healthcare, including patient harm, bias, lack of transparency, and implementation obstacles, while also proposing fourteen functional requirements for AI systems to mitigate these risks and ensure compliance with future EU regulations.
Play

DeepMind uses AI to predict harmful genetic mutations in humans

DeepMind developed an AI tool called AlphaMissense that can predict if genetic mutations in humans are harmful or not. The tool is seen as an advance in using AI to interpret genomic data, though it has limitations and does not solve everything regarding genetic mutations.
Play

Using meta-prompts to optimize performance of LLMs

Google DeepMind researchers have developed a method called Optimization by PROmpting (OPRO) to improve the performance of large language models (LLMs) like ChatGPT in solving math problems. They found that using natural language prompts, including human-style encouragement like "let's think step by step," significantly enhanced the accuracy of the AI models' results.
Play

Beats by AI

A dynamic and flowing trance music experience, set on the beach of Ibiza at 4 am, with energetic synthesizers, dramatic chords, a choir, and nostalgic and euphoric vibes.
Play

In traditional sectors, AI innovation may likely be incremental (AI)

Wealth managers are integrating AI for tasks like research and improving client interactions. While AI's use is expanding, it won't fully replace human judgment in investments. The transition is gradual, with AI tools enhancing efficiency and customer engagement, rather than causing a drastic shift in the industry.
Play

Should we be hiring AI agents vs buying AI software? (AI)

In the age of AI, there's an opportunity for startups to change the paradigm of software sales, emphasizing the benefits of selling AI-generated outcomes or work instead of merely tools or platforms.
Play

Another Brick in the Wall (Neuroscience)

Researchers at the University of California, Berkeley have reconstructed the Pink Floyd song using brain activity recordings from patients. The study aims to develop devices that vocalize thoughts, potentially helping patients who've lost the ability to speak due to brain injuries communicate naturally.
Play

This code-assistant seems to be picking up steam (AI)

Cursor is an AI-powered code assistant that can save engineers time and effort by providing quick access to documentation, code definitions, and files, as well as the ability to change entire methods or classes with a single prompt and generate code from scratch.
Play

Will generative "Everything Apps" be a thing? (AI)

New apps leveraging GenerativeAI to modify and personalize UX and features are launching. Natural, by https://brain.ai integrates with existing services you already use "Users can simply say what they want and the right app forms around their words. You no longer go to apps – apps come to you."
Play

Run automated research retrieval tasks locally (AI)

Explore how a fairly simple LLM powered retriever is being used to re-imagine web research, with the ability to be configured for privacy and observability.
Play

*OpenAI* has shut down its AI text-detection tool over inaccuracies

OpenAI has recently taken down their AI text-detection tool due to its "low rate of accuracy," and is now working on an improved version to better detect false claims generated by AI and help prevent malicious actors from using AI in their influence operations.
Play

LLM giants are forming the Frontier Model Forum for safe and responsible AI development

Today, Google, Microsoft, OpenAI and Anthropic have formed the Frontier Model Forum, a new industry body focused on ensuring safe and responsible development of frontier AI models and advancing AI safety research to benefit the entire AI ecosystem.
Play

Virtual room styling is about to get a lot better

Transform your living space with just a few clicks by using the latest technology to design on any device and shop for new furniture from the comfort of your own home!
Play

Robots + LLMs are a thing now

A new approach to 3D scene understanding that fuses multi-view images and volumetric representations to achieve state-of-the-art performance.
Play

Amazon introduces AWS Healthscribe

Amazon is expanding its range of health-focused apps and services with AWS HealthScribe, an AI-powered platform to help clinicians transcribe and analyze their conversations with patients, and AWS HealthImaging, a service designed to facilitate petabyte-scale medical imaging storage. Discover how these innovative tools can help reduce the burden of clinical documentation and improve the consultation experience.
Play

Stability AI released SDXL 1.0

Discover how SDXL 1.0 can revolutionize your artwork with its capability to generate high-quality images in virtually any art style, with vibrant and accurate colors, better contrast, lighting, and shadows, and even difficult concepts such as hands and text!
Play

New app currently in development that builds custom research agents, queries the web and synthesizes research reports for you. (AI)

GPT Researcher is an autonomous agent designed to help you quickly and accurately generate comprehensive research reports on a variety of tasks, complete with customization options for focusing on relevant resources, outlines, and lessons. Leverage the power of AI to empower your organization with accurate, unbiased, and factual information.
Play

Text-to-video and Image-to-Video are gaining momentum. (AI)

AI Tooling to generate videos, images, expand images, reimagine images, train custom models, erase things from videos, slow mo any video, make images move, remove any background and more.
Play

Enterprise specific agents based on Cohere's Command Model (AI, Enterprise)

Learn how to customize Coral to your unique job functions and connect data sources to augment its knowledge base, all while keeping data secure and private with Cohere's latest Command model.
Play

Generating high quality images from EEG with DreamDiffusion (AI, Biotech)

DreamDiffusion is a novel method for generating high-quality images directly from brain electroencephalogram (EEG) signals without the need to translate thoughts into text.
Play

New research showing AI focused developer tools could boost GDP by over $1.5 trillion (AI, Software Engineering)

GitHub Copilot has been activated by more than one million developers and adopted by over 20,000 organizations, resulting in an economic impact of $1.5 trillion and increased productivity of up to 30%, with developers feeling more fulfilled and completing tasks 55% faster.
Play

Ethical Concerns Surrounding AI Detectors in Academic Integrity

Stanford scholars have found that AI detectors used to flag cheating, plagiarism, and mis- or disinformation are not particularly reliable, especially when the real author is not a native English speaker, and can be easily subverted, raising ethical concerns about the objectivity of AI detectors and the potential for foreign-born students and workers to be unfairly accused or penalized.
Play

Gen AI may allow some to practice at the top of their license (AI)

This article provides an overview of the recent advances in generative A.I. technology that is being used by doctors to reduce the time spent on digital paperwork, freeing up their evenings and improving the patient experience.
Play

Hickok Cole uses ChatGPT to design 24-storey mixed-use building (AI, Architecture)

Explore how Hickok Cole Architectural studio has used AI chatbot ChatGPT to design a large mixed-use building with a green roof and swimming pool as a research project, and the potential of AI tools to assist the design process.
Play

Good compilation of AI related studies on the workforce (AI, Future of Work)

Generative AI is quickly making AI accessible for businesses of all sizes, with studies showing that productivity gains of up to 50% on certain knowledge-based tasks can be achieved. Learn more about how LLMs can unlock economic value and growth and how businesses can benefit from AI-powered workflows.
Play

The new partnership between Khan Academy and OpenAI seems promising (AI)

Stanford scholars have found that AI detectors used to flag cheating, plagiarism, and mis- or disinformation are not particularly reliable, especially when the real author is not a native English speaker, and can be easily subverted, raising ethical concerns about the objectivity of AI detectors and the potential for foreign-born students and workers to be unfairly accused or penalized.
Play

A new large language model called phi-1, specifically designed for code (AI)

Experience the power of phi-1 - a new large language model for code with 1.3B parameters, trained for 4 days on 8 A100s, and an accuracy of 50.6% on HumanEval - and explore its fascinating emergent properties today!
Play

Using role play for generative writing (AI)

Learn how to use Reword to make your content workflow faster and easier with this quick guide!
Play
No items found.

It takes a small village to build a great company. Let’s collab.

Office Hours  • 9am - 5pm Monday-Friday
Office Hours  • 9am - 5pm Monday-Friday
Office Hours  • 9am - 5pm Monday-Friday
Office Hours  • 9am - 5pm Monday-Friday
Office Hours  • 9am - 5pm Monday-Friday
Office Hours  • 9am - 5pm Monday-Friday
Office Hours  • 9am - 5pm Monday-Friday

Lorem ipsum dolor sit amet consectetur. Pellentesque ornare nam eu purus tortor elementum vulputate penatibus.

A beautifully minimalist note-taking app designed to mirror the way you think.