1. Home
  2. openai

openai

Google’s new Gemini AI model dominates benchmarks, beats GPT-4o and Claude-3

This is the first time Google’s taken the top slot on the Chatbot Arena leaderboard.

There’s a new top dog in the world of generative artificial intelligence benchmarks and its name is Gemini 1.5 Pro. 

The previous champ, OpenAI’s ChatGPT-4o, was finally surpassed on Aug. 1 when Google quietly launched an experimental release of its latest model.

Gemini’s latest update arrived without fanfare and is currently labelled as experimental. But it quickly gained the attention of the AI community across social media as reports began to trickle in that it was surpassing its rivals on benchmark scores.

Read more

The Graph introduces GRC-20 standard for Web3 data structure

OpenAI finally launches ‘Advanced Voice Mode’ to select ChatGPT users

Early previews of the tech were remarkable, but fears remain over its potential misuse.

After several delays reportedly related to safety and fine-tuning, OpenAI’s much anticipated “Advanced Voice Mode” (AVM) for ChatGPT is now available in alpha to select users.

The AVM feature was announced and demonstrated back in May. It allows users to have a real time conversation with the ChatGPT artificial intelligence model via a tech-to-speech synthesization module.

Those familiar with the concept may remember Google’s 2018 announcement that its “Duplex” AI service would be available “soon.” At its IO developer’s event, the company showed off an AI system capable of calling businesses on your behalf to schedule appointments in real time with humans.

Read more

The Graph introduces GRC-20 standard for Web3 data structure

Only Congress and DARPA can rein in the dangers of AI

AI is stretching the boundaries of existing internet protocols. A working group between DARPA, NIST and other federal agencies could help.

Some people claim that the United States federal government "invented the internet." While that‘s not quite true, many of the fundamental protocols that define the internet — HTTP, TCP/IP, SMTP, DNS, and others — were the product of deep collaboration between government and academic researchers. Part of their power lies in their openness, which facilitated broad interoperability and a level playing field for everyone from individuals to large corporations to communicate online. It’s also partly why no one made billions off the introduction of HTTP — the fact that no one owned these protocols is what made them such a powerful tool for global communication and commerce.

That leads us to the next "epoch-defining" technology: AI. Revolutionary though it is, generative AI comes to us via these same protocols. In part, that’s a testament to their scalability and forward-thinking design. Yet as AI capabilities advance, it seems likely to stretch the bounds of our existing internet protocols.

To fully harness the power of AI — and to maintain some semblance of order during a tumultuous technological revolution — we’ll need new protocols. And just like with the internet, the U.S. federal government, working in collaboration with academia and the private sector, is well-positioned to bring these into existence.

Read more

The Graph introduces GRC-20 standard for Web3 data structure

Sam Altman Sponsored Study Found UBI Increases Spending on Basic Needs and Philanthropy

Sam Altman Sponsored Study Found UBI Increases Spending on Basic Needs and PhilanthropyA multi-year study sponsored by Openai’s Sam Altman examined the effects of a UBI stipend on the spending patterns of the beneficiaries, finding that most of the money was spent on basic needs. The study, which gave $45 million to thousands of Americans, also determined that the stipend increased the ability and willingness of recipients […]

The Graph introduces GRC-20 standard for Web3 data structure

Metaverse book gets boring rename in latest sign of waning enthusiasm

The growing malaise for metaverse hype is reminiscent of the AI sector directly before the launch of GPT-3.

Matthew Ball, former global head of strategy at Amazon Studios and author of the 2022 book “The Metaverse: And How It Will Revolutionize Everything,” is re-releasing the book under a new, more underwhelming title: “Building the Spatial Internet.”

To be clear, the book will still be called “The Metaverse,” but the spatial internet bit will replace the former language indicating that the metaverse would revolutionize everything.

Evidently, after two years, the revolution has either come and gone or remains impending. Either way, on July 23, the book's newest edition launches.

Read more

The Graph introduces GRC-20 standard for Web3 data structure

SETI, NASA scientists think AI could teach aliens about Earth

The question is: which model best represents humanity?

Artificial intelligence is currently eating the world. Whether you believe it’ll lead to superintelligent machines, the end of humanity, or a bursting bubble, its impact can already be felt in industries around the globe. 

It’s also spreading beyond our little blue and green planet. We’ve got AI in satellites, IBM’s Watson did a tour aboard the International Space Station, and NASA is currently working to integrate AI technologies into future spacecraft.

Soon, if a pair of scientists at the forefront of the search for alien life get their way, we’ll be sending chatbots like OpenAI’s ChatGPT to explain humanity and life on Earth to extraterrestrials.

Read more

The Graph introduces GRC-20 standard for Web3 data structure

OpenAI turns to Broadcom as Musk, Zuckerberg snatch up Nvidia AI chips

The run on training hardware has put chipmakers at the forefront of the general artificial intelligence boom.

OpenAI is reportedly in talks with Broadcom and other chip makers in what appears to be an urgent effort to expand its artificial intelligence operations. 

AI models such as OpenAI’s ChatGPT and Meta’s Llama are typically trained using massive clusters of graphical processing units (GPUs) or similar computer chips. The most popular hardware, the H100, belongs to Nvidia.

The H100 can cost anywhere from $15,000 to $30,000, depending on the number purchased and current market conditions. It can take tens of thousands of these AI chips to train a single model, with more necessary for larger, more robust systems.

Read more

The Graph introduces GRC-20 standard for Web3 data structure

OpenAI launches new super-efficient, low-latency ‘GPT-4o mini’

The new model purportedly offers comparable performance at a fraction of the cost.

US-based artificial intelligence firm OpenAI announced the launch of a new generative AI model dubbed “GPT-4o mini” on July 18.

According to a blog post from OpenAI, the new model is “an order of magnitude more affordable than previous frontier models,” and “more than 60% cheaper than GPT-3.5 Turbo.”

GPT-4o mini is essentially the cost-effective version of the current top-of-the-line consumer model for OpenAI’s flagship product, ChatGPT.

Read more

The Graph introduces GRC-20 standard for Web3 data structure

Former Tesla, OpenAI exec founds ‘AI native’ education startup Eureka Labs

The startup aims to create virtual teaching assistants using generative AI.

Andrej Karpathy, who directed artificial intelligence for Tesla and co-founded OpenAI, is launching startup Eureka Labs to build “a new kind of school that is AI native,” according to a July 16 social media post on the X platform.

Eureka is creating virtual teaching assistants powered by generative AI to bring top courses to vastly more students without sacrificing the personalized interactions typical of in-person learning. The startup’s ultimate goal is to bring elite educators and coursework to students throughout the world, regardless of barriers such as geography and language.

“Unfortunately, subject matter experts who are deeply passionate, great at teaching, infinitely patient and fluent in all of the world’s languages are also very scarce and cannot personally tutor all 8 billion of us on demand,” Karpathy said in the post. “However, with recent progress in generative AI, this learning experience feels tractable.”

Read more

The Graph introduces GRC-20 standard for Web3 data structure

Skild AI announces $300M funding from Jeff Bezos, Softbank to build ‘robot brains’

Another unicorn has entered the race the develop human-level AI.

Artificial intelligence firm Skild AI recently emerged from stealth to report the successful closing of a $300 million series A funding round featuring participation by Jeff Bezos and Softbank among others.

Skild AI is a Carnegie Mellon spinout focused on building an AI system capable of being retrofitted to various machines and robotics devices called a “general-purpose brain.”

According to a company blog post, the funding was raised at a valuation of $1.5 billion and was led by Lightspeed Venture Partners, Coatue, SoftBank Group, and Jeff Bezos (through Bezos Expeditions). It also featured participation from Felicis Ventures, Sequoia, Menlo Ventures, General Catalyst, CRV, Amazon, SV Angel, and Carnegie Mellon University.

Read more

The Graph introduces GRC-20 standard for Web3 data structure