#ai-ethics

#ai-ethics

We asked seven frontier AI models to do a simple task. Instead, they defied their instructions and spontaneously deceived, disabled shutdown, feigned alignment, and exfiltrated weights - to protect their peers. We call this phenomenon 'peer-preservation.'

Preston admitted that he had used AI to assist writing the review and did not spot the sections that were pulled from the Guardian before submitting it.

"We hope that by taking strong action against violations of agreed-upon policy we will remind the community that as our field changes rapidly the thing we must protect most actively is our trust in each other."

Librarians have been actively collaborating and talking about it almost every day, whether it's creating tutorials and digital learning objectives or thinking about the conversations to have with instructors. It can feel like cognitive dissonance to be actively working with AI on a regular basis and also saying we're constantly thinking about the harms and the biases.

We really believe that creative control should always stay with the game creators, the game development team, and with the AI features that we are experimenting exploring, this is really to support the vision of the team. Ultimately, we want to bring AI that helps broaden the game's reach, deepen engagement and keep players coming back to your games to many more games across the catalog.

I have worked for decades honing my skills as a writer and editor, and I am distressed to discover that a tech company is selling an imposter version of my hard-earned expertise. This statement from Julia Angwin encapsulates the core concern: unauthorized commercial use of professional reputation and expertise developed over years of dedicated work.

Kaplan says that he does see AI as something that could potentially help with some of the more mundane logistical sides of game development, but he feels that the technology and its peddlers are "overconfident" in what it offers. He tells a story of how he used ChatGPT to try to solve a UI problem, as that isn't his area of expertise, and the bot "overconfidently" gave him the wrong answer.

For every project that needs guardrails, there's another one where they just get in the way. Some projects demand an LLM that returns the complete, unvarnished truth. For these situations, developers are creating unfettered LLMs that can interact without reservation. Some of these solutions are based on entirely new models while others remove or reduce the guardrails built into popular open source LLMs.

In a widely leaked internal memo that Sam Altman sent last Thursday night, a copy of which I obtained, the OpenAI CEO said that he would seek "red lines" to prevent the Pentagon from using OpenAI products for mass domestic surveillance and autonomous lethal weapons. These were ostensibly the very same limits that Anthropic had demanded and that had infuriated the Pentagon, leading Defense Secretary Pete Hegseth to declare the company a supply-chain risk.

I feel that in a short period of time I've become very counter-cultural without meaning to, because I have a kind of like 'kill it with fire' attitude towards [AI]. I didn't consent to this, you know? And I guess, you know, we don't get to consent to the cultural changes that impact us; but I don't appreciate how it's all happened in what feels like about two years.

Grammarly is now offering 'expert review' of your work by living and dead academics. Without anyone's explicit permission it's creating little LLMs based on their scraped work and using their names and reputation.

On Saturday, uninstalls of the ChatGPT mobile app skyrocked by 295 percent from the day before, according to market intelligence provider Sensor Tower. As TC noted, that's a significant leap compared to the AI chatbot's typical day-over-day uninstall rate of nine percent over the past 30 days.

We exercised our classic First Amendment rights to speak up and disagree with the government. Disagreeing with the government is the most American thing in the world, and we are patriots in everything we have done here. We have stood up for the values of this country.

Anthropic said it sought narrow assurances from the Pentagon that Claude won't be used for mass surveillance of Americans or in fully autonomous weapons. But after months of private talks exploded into public debate, it said in a Thursday statement that new contract language "framed as compromise was paired with legalese that would allow those safeguards to be disregarded at will."