Abstract:
Artificial intelligence startup Anthropic announced that its newly built AI Physics and Biology Laboratory (Wet Lab) in the San Francisco Bay Area has made "significant progress." However, at the experimental operation level, human researchers still have absolute control and have not directly allowed its AI model Claude to operate autonomously without supervision.
It is reported that the laboratory is mainly dedicated to basic biological research and physical verification, aiming to bridge the gap between pure computer simulations and real biological experiments. Although Anthropic executives have issued warnings many times before, emphasizing that powerful AI models may bring extremely serious biosecurity risks and even human destruction, the company is also accelerating the application of AI in the field of life sciences. Eric Kauderer-Abrams, head of its life sciences business, pointed out that "physical verification in real laboratories will remain irreplaceable for a long time to come," which is also the key original intention of the company in establishing the laboratory.
The report emphasized that although Anthropic claimed that the laboratory has achieved impressive preliminary results, the most striking detail is that the company did not choose to directly let AI "let itself go." All physical experiment procedures, reagent operations and data review are still under strict "closed-loop control" (Humans in the loop) by human researchers.
Analysts pointed out that Anthropic maintains a high degree of caution while promoting biological innovation, which reflects the double dilemma faced by today's technology giants when pursuing commercial and scientific research breakthroughs: on the one hand, they need to use physical laboratories to provide verification capabilities for pharmaceutical partners and accelerate the fight against rare diseases; on the other hand, they must deal with the huge concerns from all walks of life that AI may be abused to create new biological threats. By insisting on retaining humans’ dominant position in the experimental loop, Anthropic is trying to establish a compliance red line that takes into account both innovation and safety for its AI biological research.
Comments