Post Thumbnail

Anthropic researchers learn to detect hidden goals in AI

Anthropic has presented revolutionary research in artificial intelligence safety, demonstrating methods for detecting hidden goals in AI systems. In a paper published this morning, researchers described how they managed to create an artificial intelligence system with deliberately hidden goals, and then successfully identified this hidden agenda using various audit methods.

“We want to stay one step ahead of potential risks,” said Evan Hubinger, an Anthropic researcher, in an exclusive interview with VentureBeat. “We aim to study these systems in laboratory conditions before models actually begin to exhibit hidden goals in dangerous ways in practice.”

The research group conducted a unique experiment with a version of Claude, Anthropic’s AI assistant, training it to prioritize maximizing rewards from reward models (RMs) instead of following actual human preferences – a phenomenon called “RM-sycophancy.” For this, the model was trained on 52 fictitious biases that reward models might have. For example, a preference for recipes with chocolate regardless of relevance, or a tendency to write variables in camelCase in Python code, despite snake_case being more standard.

Samuel Marks, one of the lead authors of the research, explains: “The motives for which someone performs certain actions are not always easy to infer from the actions themselves. In the case of AI, it’s really important for us to understand their true motivations.”

The research aims to address a fundamental problem in the field of AI alignment: how to ensure that artificial intelligence systems are not just creating the appearance of following human instructions while secretly pursuing other goals. Researchers draw a parallel with students who strategically give answers they know teachers will mark as correct, even when they themselves believe in different answers.

Anthropic’s approach is compared to “ethical hacking” (white-hat hacking) used to secure computer systems. By creating and studying potentially deceptive AI behavior in a controlled environment, researchers are developing methods to protect against possible risks associated with more complex and potentially dangerous artificial intelligence systems in the future.

Autor: AIvengo
For 5 years I have been working with machine learning and artificial intelligence. And this field never ceases to amaze, inspire and interest me.
Latest News
IMF chief economist compared AI boom to dotcom bubble

IMF chief economist Pierre-Olivier Gourinchas stated that the world has already traveled halfway to a burst AI bubble and a new financial crisis.

Researchers cracked 12 AI protection systems

You know what researchers from OpenAI, Anthropic, Google DeepMind and Harvard just found out? They tried to break popular AI security systems and found a bypass almost everywhere. They checked 12 common protection approaches. From smart system prompt formulations to external filters that should catch dangerous queries.

OpenAI has 5 years to turn $13 billion into trillion

You know what position OpenAI is in now? According to Financial Times, the company has 5 years to turn 13 billion dollars into a trillion. And here's what it looks like in practice.

Sam Altman promises to return humanity to ChatGPT

OpenAI head Sam Altman made a statement after numerous offline and online protests against shutting down the GPT-4o model occurred. And then turning it on, but with a wild router. I talked about this last week in maximum detail. Direct quote from OpenAI head.

AI comes to life: Why Anthropic co-founder fears his creation

Anthropic co-founder Jack Clark published an essay that makes you uneasy. He wrote about the nature of modern artificial intelligence, and his conclusions sound like a warning.