March 29 (Reuters) - Elon Musk and a group of artificial intelligence experts and industry executives are calling for a six-month pause in developing systems more powerful than OpenAI's newly launched GPT-4, in an open letter citing potential risks to society. Earlier this month, Microsoft-backed OpenAI unveiled the fourth iteration of its GPT (Generative Pre-trained Transformer) AI program, which has wowed users by engaging them in human-like conversation, composing songs and summarising lengthy documents. "Powerful AI systems should be developed only once we are confident that their effects will be positive and their risks will be manageable," said the letter issued by the Future of Life Institute. The non-profit is primarily funded by the Musk Foundation, as well as London-based group Founders Pledge, and Silicon Valley Community Foundation, according to the European Union's transparency register. "AI stresses me out," Musk said earlier this month.

Elon Musk and a group of artificial intelligence experts and industry executives are calling for a six-month pause in developing systems more powerful than OpenAI's newly launched GPT-4, in an open letter citing potential risks to society. Earlier this month, Microsoft-backed OpenAI unveiled the fourth iteration of its GPT (Generative Pre-trained Transformer) AI program, which has wowed users by engaging them in human-like conversation, composing songs and summarizing lengthy documents. "Powerful AI systems should be developed only once we are confident that their effects will be positive and their risks will be manageable," said the letter issued by the Future of Life Institute. This could be due to a conflict with your ad-blocking or security software. Please add japantimes.co.jp and piano.io to your list of allowed sites.

Goldman Sachs has said generative artificial intelligence (AI) systems might disrupt the labor market and impact 300 million full-time jobs worldwide. The emergence of generative AI, known for its ability to produce text and other content based on users' requests, is marked by its popularity. It spawned from the release of OpenAI's ChatGPT, which quickly captured users' attention and prompted other tech companies to follow suit and launch their own AI systems. After scrutinizing occupational tasks data from the U.S. and Europe, Goldman Sachs analysts, in their research report,projected that approximately 300 million job positions worldwide could be vulnerable to automation, provided generative AI delivers on its promised capabilities. Around 66% of existing jobs face the possibility of being affected by artificial intelligence automation in varying degrees, analysts Joseph Briggs and Devesh Kodnani noted in the report.

We investigate whether large language models can perform the creative hypothesis generation that human researchers regularly do. While the error rate is high, generative AI seems to be able to effectively structure vast amounts of scientific knowledge and provide interesting and testable hypotheses. The future scientific enterprise may include synergistic efforts with a swarm of "hypothesis machines", challenged by automated experimentation and adversarial peer reviews. In a university or research institute, a significant portion of fresh ideas arises out of discussions.

The unlearning problem of deep learning models, once primarily an academic concern, has become a prevalent issue in the industry. The significant advances in text-to-image generation techniques have prompted global discussions on privacy, copyright, and safety, as numerous unauthorized personal IDs, content, artistic creations, and potentially harmful materials have been learned by these models and later utilized to generate and distribute uncontrolled content. To address this challenge, we propose \textbf{Forget-Me-Not}, an efficient and low-cost solution designed to safely remove specified IDs, objects, or styles from a well-configured text-to-image model in as little as 30 seconds, without impairing its ability to generate other content. Alongside our method, we introduce the \textbf{Memorization Score (M-Score)} and \textbf{ConceptBench} to measure the models' capacity to generate general concepts, grouped into three primary categories: ID, object, and style. Using M-Score and ConceptBench, we demonstrate that Forget-Me-Not can effectively eliminate targeted concepts while maintaining the model's performance on other concepts. Furthermore, Forget-Me-Not offers two practical extensions: a) removal of potentially harmful or NSFW content, and b) enhancement of model accuracy, inclusion and diversity through \textbf{concept correction and disentanglement}. It can also be adapted as a lightweight model patch for Stable Diffusion, allowing for concept manipulation and convenient distribution. To encourage future research in this critical area and promote the development of safe and inclusive generative models, we will open-source our code and ConceptBench at \href{https://github.com/SHI-Labs/Forget-Me-Not}{https://github.com/SHI-Labs/Forget-Me-Not}.

Decoding human visual neural representations is a challenging task with great scientific significance in revealing vision-processing mechanisms and developing brain-like intelligent machines. Most existing methods are difficult to generalize to novel categories that have no corresponding neural data for training. The two main reasons are 1) the under-exploitation of the multimodal semantic knowledge underlying the neural data and 2) the small number of paired (stimuli-responses) training data. To overcome these limitations, this paper presents a generic neural decoding method called BraVL that uses multimodal learning of brain-visual-linguistic features. We focus on modeling the relationships between brain, visual and linguistic features via multimodal deep generative models. Specifically, we leverage the mixture-of-product-of-experts formulation to infer a latent code that enables a coherent joint generation of all three modalities. To learn a more consistent joint representation and improve the data efficiency in the case of limited brain activity data, we exploit both intra- and inter-modality mutual information maximization regularization terms. In particular, our BraVL model can be trained under various semi-supervised scenarios to incorporate the visual and textual features obtained from the extra categories. Finally, we construct three trimodal matching datasets, and the extensive experiments lead to some interesting conclusions and cognitive insights: 1) decoding novel visual categories from human brain activity is practically possible with good accuracy; 2) decoding models using the combination of visual and linguistic features perform much better than those using either of them alone; 3) visual perception may be accompanied by linguistic influences to represent the semantics of visual stimuli. Code and data: https://github.com/ChangdeDu/BraVL.

"How well does the AI perform clinically? And my answer is, I'm stunned to say: Better than many doctors I've observed."--Isaac The large language model GPT-4 (LLM, aka generative AI chatbot or foundation model) was just released 2 weeks ago (14 March) but there's already been much written about its advance beyond ChatGPT, released 30 November 2022, "the most successful new product in the history of the western world" with over 100 million users in just 2 months. A new book by Peter Lee, Carey Goldberg, Isaac Kohane will be released as an e-book April 15th and as a paperback May 3rd and I've had the chance to read it. With my keen interest for how AI can transform medicine (as written about in Deep Medicine and multiple recent review papers here, here, here), I couldn't put it down.

You can't scan the headlines lately without seeing buzz around generative artificial intelligence (AI). The product innovations are only beginning. But even with the best technology out there, you'll still be faced with a key question: How can you implement AI at scale in a way that maximizes the return on your investment? Let's look at one model company you can learn from. Schneider Electric, a global energy management and industrial automation company, has formalized an AI program under a new Chief AI Officer and scaled it to every corner of the company.

Generative AI is the latest big tech trend, with the latest variations of text and image generators now able to create original content that's comparable to human outputs, opening up a range of new possibilities. That's also freaking a lot of people out, due to concerns that they could be out of the job entirely due to the sudden influx of impressive AI tools. And some, like digital artists, are already feeling the pinch – but it is worth noting that AI systems can only iterate on what's come before, in order to provide similar content, they can't come up with entirely original, unique, or even trustworthy material. 'Trustworthy' in this context relates to the accuracy of the text data such systems provide, with AI systems known to'hallucinate' answers based on the various data points they can connect to your query. Essentially, you really have to know and understand the topics that you're focusing on to produce the best results, because you can then view the outputs with a more critical eye, and ensure no mistakes or errors slip through.

Experts have fretted about the risks of building supersmart AIs for years, but the conversation has become louder over the last six months as new image generators and chatbots that can have eerily humanlike conversations have been released to the public. Interacting with the newly-released chatbots like OpenAI's GPT4 has prompted many to declare that a human-level AI is just around the corner, but other experts point out that the way the chatbots work is by simply guessing the right words to say next based on their training, which included reading trillions of words online. The bots often devolve into bizarre conversational loops if prompted for long enough, and pass off made-up information as factual.