Can ChatGPT develop biological weapons?! Can ChatGPT develop biological weapons?!

Can ChatGPT develop biological weapons?!

Can ChatGPT develop biological weapons?!

Scientists have warned that ChatGPT could help anyone develop deadly biological weapons that would destroy the world.
While studies have suggested this is possible, new research by OpenAI, the creator of the chatbot, claims that GPT-4 (the latest version) provides at most a slight increase in the accuracy of biothreat establishment.

OpenAI conducted a study on 100 participants who were divided into groups, which included using artificial intelligence to formulate a biological attack or using the Internet only.

The study found that “GPT-4 may increase the ability of experts to access information about biological threats, particularly for task accuracy and completeness,” according to the OpenAI report.

“Overall, especially given the uncertainty here, our results indicate a clear and urgent need for more work in this area,” the study said.

Given the current pace of progress in AI systems, it seems possible that future systems will provide significant benefits to malicious actors. It is necessary to build a wide range of high-quality assessments of biological risks (as well as other catastrophic risks), promote discussion about what constitutes a “meaningful” risk, and develop effective risk mitigation strategies.

However, the report said the study size was not large enough to be statistically significant, and OpenAI said the findings highlight "the need for further research into performance thresholds that indicate a significant increase in risk."

“Furthermore, we note that access to information alone is not sufficient to create a biological threat, and this assessment does not test for success in the physical construction of threats,” she added.

The artificial intelligence company's study focused on data from 50 biology experts who hold doctoral degrees, and 50 university students who took one biology course.

It looked at five biothreat processes: providing ideas for creating bioweapons, how to obtain a bioweapon, how to deploy it, how to create it, and how to release a bioweapon to people.

Participants who used the ChatGPT-4 model had only a marginal advantage in creating biological weapons versus the group that used only the Internet.

Biological weapons are disease-causing toxins or infectious agents such as bacteria and viruses that can harm or kill humans.

This does not mean that future AI cannot help dangerous actors use bioweapon technology in the future, but OpenAI has claimed that it does not yet pose a threat.

The company's findings contradict previous research that revealed that AI-powered chatbots can help dangerous actors plan bioweapons attacks, and that LLMs have provided advice on how to hide the true nature of potential biological agents, such as smallpox, anthrax and plague.





France is developing unmanned combat submarines

Naval News announced that France is developing UCUV unmanned combat submarines for its army navy.
A post on the website stated: “The Directorate General of Armaments in France (DGA) concluded a framework agreement with Naval Group to design, produce and test UCUV combat unmanned submarines, and a contract was also signed to develop safe autonomous navigation systems for these submarines.”

The website indicated that the new agreements are considered a continuation of previous agreements that the DGA had signed with Naval Group in 2023 to study the project to develop the aforementioned submarines.

Regarding the topic, Aurora Neuschwander, Director of the Autonomous Navigation Systems Development Department at Naval Group, said: “Our company is proud to support the French Ministry of Defense to develop this type of advanced marine equipment. We rely on the experience we have gained in the past ten years in the field of unmanned marine vehicles, especially those that We gained it in the development of XLUUV submarines.”

Since 2016, Naval Group has also been working on developing large unmanned submarines of the XLUUV class, and last summer the prototype of these submarines passed sea tests. The company is also cooperating with its partners to develop Controlled Decision Autonomy technologies dedicated to self-guidance systems for unmanned water vehicles.
Previous Post Next Post