#ai-safety

#ai-safety

This process, becoming aware of something not working and then changing what you're doing, is the essence of metacognition, or thinking about thinking. It's your brain monitoring its own thinking, recognizing a problem, and controlling or adjusting your approach. In fact, metacognition is fundamental to human intelligence and, until recently, has been understudied in artificial intelligence systems. My colleagues Charles Courchaine, Hefei Qiu, Joshua Iacoboni, and I are working to change that.

Humanity is entering a phase of artificial intelligence development that will test who we are as a species, the boss of leading AI startup Anthropic has said, arguing that the world needs to wake up to the risks. Dario Amodei, co-founder and chief executive of the company behind the hit chatbot Claude, voiced his fears in a 19,000-word essay entitled the adolescence of technology. Describing the arrival of highly powerful AI systems as potentially imminent, he wrote:

Meta is pressing ahead with the global monetisation of Threads, confirming plans to roll out advertising to users worldwide. The expansion will be phased over several months, starting this week, as the company seeks to balance revenue growth with user experience. Brands will be able to run image and video formats, including carousel ads and the newer 4:5 aspect ratio, and manage campaigns alongside Facebook, Instagram and WhatsApp through Meta's Business Settings.

Microsoft wasted little time last fall after reaching a deal to finalize its new relationship with OpenAI to find a new AI dance partner - Anthropic, the second most valuable AI startup in the world. Even though the relationship between Microsoft and Anthropic is only a few months old, it appears as if Microsoft sees a future with Anthropic that's at least as valuable as the one it had with OpenAI.

Anthropic has published a new constitution for its AI model Claude. In this document, the company describes the values, behavioral principles, and considerations that the model must follow when processing user questions. The constitution has been made publicly available under a Creative Commons CC0 license, allowing the content to be used freely without permission. Anthropic published the first version of this constitution in May 2023.

Hinton, who helped pioneer the neural networks that underpin modern artificial intelligence, has become one of the field's most outspoken critics as AI systems grow more powerful and widespread. He has predicted that AI could trigger widespread job losses, fuel social unrest, and eventually outsmart humans - and has said that researchers should focus more on how advanced systems are trained, including ensuring they are designed to protect human interests.

Here is a recap of what happened in the search forums today, through the eyes of the Search Engine Roundtable and other search forums on the web. OpenAI will be testing ads in ChatGPT very soon. Google's Gemini 3 Pro now powers some AI Overviews. Surprise, surprise, Google is appealing the search monopoly ruling. Google warns that using free subdomian hosts is not a good idea. Google also said that comment link spam won't help or hurt your site.

You could say it all started with Elon Musk's AI FOMO - and his crusade against "wokeness." When his AI company, xAI, announced Grok in November 2023, it was described as a chatbot with "a rebellious streak" and the ability to "answer spicy questions that are rejected by most other AI systems." The chatbot debuted after a few months of development and just two months of training, and the announcement highlighted that Grok would have real-time knowledge of the X platform.

OpenAI's new LLM has revolutionized AI and opened up new possibilities for marketers. Here's a look at how three big-name brands have embraced the technology. In March, the AI lab OpenAI released GPT-4, the latest version of the large language model (LLM) behind the viral chatbot ChatGPT. Since then, a small number of brands have been stepping forward to integrate the new-and-improved chatbot into their product development or marketing efforts. To a certain extent, this has required some courage.

AI labs just can't get their employees to stay put. Yesterday's big AI news was the abrupt and seemingly acrimonious departure of three top executives at Mira Murati's Thinking Machines lab. All three were quickly snapped up by OpenAI, and now it seems they won't be the last to leave. Alex Heath is reporting that two more employees are expected to leave for OpenAI in the next few weeks.

"We have implemented technological measures to prevent the Grok account from allowing the editing of images of real people in revealing clothing such as bikinis," the company's safety team said in a statement, adding that the restrictions applied to all users, including paid subscribers. "We now geoblock the ability of all users to generate images of real people in bikinis, underwear, and similar attire via the Grok account and in Grok in X in those jurisdictions where it's illegal," the statement said.

Built into the Claude Desktop app, the new tool lets users designate a specific folder where Claude can read or modify files, with further instructions given through the standard chat interface. The result is similar to a sandboxed instance of Claude Code, but requires far less technical savvy to set up. Currently in research preview, Cowork is only available to Max subscribers, with a waitlist available for users on other plans.

From reproductive rights to climate change to Big Tech, The Independent is on the ground when the story is developing. Whether it's investigating the financials of Elon Musk's pro-Trump PAC or producing our latest documentary, 'The A Word', which shines a light on the American women fighting for reproductive rights, we know how important it is to parse out the facts from the messaging.

Whether it's investigating the financials of Elon Musk's pro-Trump PAC or producing our latest documentary, 'The A Word', which shines a light on the American women fighting for reproductive rights, we know how important it is to parse out the facts from the messaging. At such a critical moment in US history, we need reporters on the ground. Your donation

OpenAI is launching a new facet for its AI chatbot called ChatGPT Health. This new feature will allow users to connect medical records and wellness apps to ChatGPT in order to get more tailored responses to queries about their health. The company noted that there will be additional privacy safeguards for this separate space within ChatGPT, and said that it will not use conversations held in Health for training foundational models. ChatGPT Health is currently in a testing stage, and there are some regional restrictions on which health apps can be connected to the AI company's platform.

Late last week, Elon Musk's Grok chatbot unleashed a flood of images of women, nude and in very little clothing, both real and imagined, in response to users' public requests on X, formerly Twitter. Mixed in with the generated images of adults were ones of young girls children likewise wearing minimal clothing, according to Grok itself. In an unprecedented move, the chatbot itself apologized while its maker, xAI, remained silent:

how many grams of kratom gets you a strong high?

How many grams of kratom gets you a strong high?

If we're to believe Yoshua Bengio, one of the so-called "godfathers" of AI, some advanced models are showing signs of self-preservation - which is exactly why we shouldn't endow them with any kind of rights whatsoever. Because if we do, he says, theymay run away with that autonomy and turn on us before we have a chance to pull the plug. Then it's curtains for this whole "humankind" experiment.