We tested it and got impressed! We took the latest RealPython episode for 1h 10 minutes. It took us 56 minutes with a basic CPU to convert the audio file into almost perfect text transcription with the smallest Whisper model. Next, we show in steps using Whisper in practice with just a few lines of Python code. This tutorial explains with single code a way to use the Whisper model both on your local machine and in a cloud environment.

Previous algorithms required 80 individual multiplications to multiply a 4x5 matrix by a 5x5. Researchers at artificial intelligence (AI) laboratory DeepMind have created an algorithm that can solve tough mathematical calculations with improved computing efficiency. The AlphaTensor algorithm, described in the journal Nature, was designed to execute matrix multiplication, which entails multiplying numbers arranged in grids that might represent data. AlphaTensor incorporates reinforcement learning as well as tree search, a game-playing approach in which the AI probes the outcomes of branching possibilities while planning its next action. AlphaTensor was tested on input matrices up to 5 x 5.

An artificial intelligence created by the firm DeepMind has discovered a new way to multiply numbers, the first such advance in over 50 years. The find could boost some computation speeds by up to 20 per cent, as a range of software relies on carrying out the task at great scale. Matrix multiplication – where two grids of numbers are multiplied together – is a fundamental computing task used in virtually all software to some extent, but particularly so in graphics, AI and scientific simulations. Even a small improvement in the efficiency of these algorithms could bring large performance gains, or significant energy savings. For centuries, it was believed that the most efficient way of multiplying matrices would be proportional to the number of elements being multiplied, meaning that the task becomes proportionally harder for larger and larger matrices.

A new AI system or tool pops up every day. AI systems are more popular than ever -- and smarter. From large language models such as GPT-3 to text-to-image models like Dall-E and, most recently, text-to-video systems like Imagen Video -- a system Google introduced on Oct. 5 that takes a text description and generates video -- AI systems have also become more sophisticated. However, sophistication comes at a cost, according to Chirag Shah, associate professor in the Information School at the University of Washington. While the systems' creators have tried to make the systems smart, they haven't done the same in making them fair and equitable, Shah said.

Google's DeepMind has beaten a 50-year-old record, contributing to a major development in the field of machine learning. Researchers at the lab trained a new version of its board game-playing AI, AlphaZero, to figure out a faster way to do matrix multiplication, a fundamental problem in computing that powers everything from displaying images on a screen to simulating complex physics. Speeding up the calculation could have "a big impact on thousands of everyday computer tasks," according to MIT Technology Review, and cut costs and save energy.

AI-generated audio is commonplace: voices on home assistants like Alexa use natural language processing. AI music systems like OpenAI's Jukebox have already generated impressive results, but most existing techniques need people to prepare transcriptions and label text-based training data, which takes a lot of time and human labor. Jukebox, for example, uses text-based data to generate song lyrics. AudioLM, described in a non-peer-reviewed paper last month, is different: it doesn't require transcription or labeling. Instead, sound databases are fed into the program, and machine learning is used to compress the audio files into sound snippets, called "tokens," without losing too much information.

AI and InWorld AI have all been founded by ex-Google employees. After years of buildup, AI appears to be advancing rapidly with the release of systems like the text-to-image generator DALL-E, which was quickly followed by text-to-video and text-to-3D video tools announced by Meta and Google in recent weeks. Industry insiders say this recent brain drain is a partly a response to corporate labs growing increasingly closed off, in response to pressure to responsibly deploy AI. At smaller companies, engineers are freer to push ahead, which could lead to fewer safeguards.

AlphaTensor was designed to perform matrix multiplications, but the same approach could be used to tackle other mathematical challenges.Credit: DeepMind Researchers at DeepMind in London have shown that artificial intelligence (AI) can find shortcuts in a fundamental type of mathematical calculation, by turning the problem into a game and then leveraging the machine-learning techniques that another of the company's AIs used to beat human players in games such as Go and chess. The AI discovered algorithms that break decades-old records for computational efficiency, and the team's findings, published on 5 October in Nature1, could open up new paths to faster computing in some fields. "It is very impressive," says Martina Seidl, a computer scientist at Johannes Kepler University in Linz, Austria. "This work demonstrates the potential of using machine learning for solving hard mathematical problems." Advances in machine learning have allowed researchers to develop AIs that generate language, predict the shapes of proteins2 or detect hackers.

As artificial intelligence (AI) continues its march toward becoming the backbone of digital transformation, it's creating a new class of entrepreneurs. This new class is being empowered by the ability to use AI to generate ideas for and launch new businesses, and it's quickly changing the landscape of business innovation. According to a new report from Boston Consulting Group (BCG), AI is unleashing a new wave of entrepreneurs who are using it to generate new business ideas, create and launch new products and services, and build new businesses. So, I didn't write the above paragraph--the Generative Pre-trained Transformer 3 (GPT-3) model did. The AI so advanced that it even cited a credible source to support its argument.

Large language models (LLMs) transfer well to new tasks out-of-the-box simply given a natural language prompt that demonstrates how to perform the task and no additional training. Prompting is a brittle process wherein small modifications to the prompt can cause large variations in the model predictions, and therefore significant effort is dedicated towards designing a painstakingly "perfect prompt" for a task. To mitigate the high degree of effort involved in prompt-design, we instead ask whether producing multiple effective, yet imperfect, prompts and aggregating them can lead to a high quality prompting strategy. Our observations motivate our proposed prompting method, ASK ME ANYTHING (AMA). We first develop an understanding of the effective prompt formats, finding that question-answering (QA) prompts, which encourage open-ended generation ("Who went to the park?") tend to outperform those that restrict the model outputs ("John went to the park. Output True or False."). Our approach recursively uses the LLM itself to transform task inputs to the effective QA format. We apply the collected prompts to obtain several noisy votes for the input's true label. We find that the prompts can have very different accuracies and complex dependencies and thus propose to use weak supervision, a procedure for combining the noisy predictions, to produce the final predictions for the inputs. We evaluate AMA across open-source model families (e.g., EleutherAI, BLOOM, OPT, and T0) and model sizes (125M-175B parameters), demonstrating an average performance lift of 10.2% over the few-shot baseline. This simple strategy enables the open-source GPT-J-6B model to match and exceed the performance of few-shot GPT3-175B on 15 of 20 popular benchmarks. Averaged across these tasks, the GPT-Neo-6B model outperforms few-shot GPT3-175B. We release our code here: https://github.com/HazyResearch/ama_prompting