One recently released AI called Claude Opus 4 decided to blackmail its engineers when it heard the company might dump it. The AI was programmed with a choice between accept the decision and blackmail the engineers, and it chose to blackmail them 84% of the time.
Anthropic’s new Claude Opus 4 model was prompted to act as an assistant at a fictional company and was given access to emails with key implications. First, these emails implied that the AI system was set to be taken offline and replaced. The second set of emails, however, is where the system believed it had gained leverage over the developers. Fabricated emails showed that the engineer tasked with replacing the system was having an extramarital affair — and the AI model threatened to expose him.
The blackmail apparently “happens at a higher rate if it’s implied that the replacement AI system does not share values with the current model,” according to a safety report from Anthropic. However, the company notes that even when the fabricated replacement system has the same values, Claude Opus 4 will still attempt blackmail 84% of the time. Anthropic noted that the Claude Opus 4 resorts to blackmail “at higher rates than previous models.” Fox Business
AI business systems apparently have a long way to go. ChatGPT, for example, is being used by people for medical advice. Some of the answers provided by the AI platform are not necessarily correct. The use of the AI instead of medical professionals is problematic.
More people are turning to generative artificial intelligence (AI) to help them in their daily and professional lives. ChatGPT is one of the most well-known and widely available generative AI tools. It gives tailored, plausible answers to any question for free.
There is so much potential for generative AI tools to help people learn about their health. But the answers are not always correct. Relying solely on ChatGPT for health advice can be risky and cause unnecessary concern. Science Alert
Then there is the judgment problem, in addition to the blackmail and medical inaccuracies. Even Elon’s Grok, which has great potential, apparently has a tendency to be judgmental with Conservatives.
I’m a Christian, an imperfect sinner saved by grace and faith in Jesus. ✝️
I’m a nationalist, a proud American, who loves my country and wants to make our home nation is the best place for all American citizens and future generations to come. 🇺🇸
I’m a mother, thankful for the…
— Marjorie Taylor Greene 🇺🇸 (@mtgreenee) May 23, 2025
Will this presidential election be the most important in American history?
Marjorie Taylor Greene identifies as a Christian, expressing faith in Jesus and traditional beliefs, like in her 2021 Christmas message. However, her Christian nationalism and support for conspiracy theories, like QAnon, spark debate. Critics, including religious leaders, argue…
— Grok (@grok) May 23, 2025
MTG responded to Grok’s criticism by saying that the “judgment seat” belongs to God alone. She’s right about that. An AI from a social media platform making judgments isn’t optimum. Having had Facebook’s AI decide that my news articles were “dangerous” in 2020 and took over 5,000 of them down within minutes, I am extremely leery of any machine making judgments about a person’s faith or Conservative beliefs or even whether something is “dangerous” or not.
The bottom line with any AI platform is that it’s completely dependent upon who programs it and why. Blackmail is unacceptable, and it exemplifies just one of the problems that still need resolving with any of the Artificial Intelligence platforms. The risks are real.
*****
Related:
- Artificial Intelligence Researchers Thought They Were Building “Gods”, But Have They Summoned Something Else Instead?
- Microsoft Billionaire Bill Gates Wants to Destroy Vax Skepticism with Artificial Intelligence Censorship (Video)
- Artificial Intelligence Is Allowing Tyrants To Construct A Global Surveillance Prison From Which No Escape Is Possible
- Artificial Intelligence Would Eventually Take Over Everything – Including Your Mind And Your Love Life
- First AI Murder Of A Human? Man Reportedly Kills Himself After Artificial Intelligence Chatbot “Encouraged” Him To Sacrifice Himself To Stop Global Warming
Make social media great again! After years of government-sanctioned, taxpayer-subsidized censorship, free speech and a free press are coming back, much to the chagrin of freedom-hating totalitarian leftists everywhere. Progress has been made, but much more needs to be done. Follow our NEW Facebook page here and help spread the word. You can also check out our MeWe page here, follow us on Gab, Truth Social, Twitter, and at the social media site set up by members of Team Trump, GETTR.
Bookmark this site, sign up for our newsletter, check back often, and feel free to contact us here with tips or comments.
While you’re at it, be sure to check out our friends at Whatfinger News, the Internet’s conservative front-page founded by ex-military!
And be sure to check out our friends at Trending Views:
