Accusations of Reckless Development
A group of AI whistleblowers have written an open letter to tech giants like Google and Open AI, accusing them of a reckless race to develop technology that could endanger humanity.
Risks and Lack of Oversight
According to a letter signed by current and former employees of OpenAI, Google DeepMind, and Anthropic, AI companies are motivated by financial reasons to avoid oversight and point to the absence of federal regulations on advanced AI development.
Employees highlight possible dangers such as the dissemination of false information, deepening disparities, and the potential relinquishment of authority over self-governing AI systems, which could lead to human annihilation.
Call to Action for Transparency and Protection
The letter asks AI giants to commit to four principles designed to boost transparency and protect whistleblowers who speak out publicly.
ChatGPT Outage Affects Millions
In other news, ChatGPT experienced a “major outage” that affected millions of users worldwide. OpenAI reported the issue on June 4 and pushed out a fix, but many users are still experiencing problems. This is the second major outage of the day, with the first lasting around 90 minutes.
Previous Outages and Issues of Open AI
ChatGPT service is experiencing its second major outage of the day, with many users not seeing responses to queries. The service was down for around 90 minutes in November, impacting OpenAI’s API services.
The second round of issues started at 10:30AM ET, with thousands of reports on Downdetector. OpenAI has confirmed that the November outage was related to a DDoS attack. A Microsoft outage also took down ChatGPT search features last month.
Ongoing Efforts to Resolve the Issue
OpenAI is currently investigating the issue and hopes that the fix doesn’t last the five or so hours that it took to fix the last outage. OpenAI has reported a “major outage” for ChatGPT, which has returned from its sick leave.
The outage, which lasted just over two hours, has been reported by Downdetector and is currently being monitored by OpenAI. The last serious outage was in May 2023, when the service was down for around eight hours. The chatbot has not been immune to outages, but this is the biggest and longest one of the year.
The issues are mostly affecting logged-in users, as OpenAI is still rolling out the option of using the service instantly without logging in. However, the problems are slowly getting resolved, and users can now access the latest GPT-4o model on their Plus account and in the mobile app.
Downdetector is also seeing a drop in reports following a second spike earlier. The chatbot has been criticized for its confusing and threatening responses from users. The next step is for OpenAI to provide a more promising status update.
Lucky Khan specializes in event and incident coverage, bringing his keen insights and comprehensive reporting to the forefront. With a knack for capturing the essence of major happenings, Lucky keeps his readers informed and engaged with the latest updates.