Source: Fox News
As tech experts warn that the rapid evolution of artificial intelligence could threaten humanity, OpenAI's ChatGPT weighed in with its own predictions on how humanity could be wiped off the face of the Earth.
Fox News Digital asked the chatbot to weigh in on the apocalypse, and it shared four possible scenarios of how humanity could ultimately be wiped out.
"It's important to note that predicting the end of the world is a difficult and highly speculative task, and any predictions in this regard should be viewed with skepticism," the bot responded. "However, there are several trends and potential developments that could significantly impact the trajectory of humanity and potentially contribute to its downfall."
Fears that AI could spell the end of humanity has for years been fodder for fiction but has become a legitimate talking point among experts as tech rapidly evolves – with British theoretical physicist Stephen Hawking issuing a dire warning back in 2014
"The development of full artificial intelligence could spell the end of the human race," he said then. Hawking died in 2018.
The sentiment has only intensified among some experts nearly a decade later, with tech giant Elon Musk saying this year that the tech "has the potential of civilizational destruction."
ChatGPT kicked off its grim predictions by speculating that climate change will have "catastrophic effects on the planet if not addressed," pointing to how rising sea levels, extreme weather and food scarcity could lead to "widespread displacement, conflict, and instability."
When asked in a follow-up question exactly how climate change could end the world, the chatbot elaborated that coastal flooding, displacement of millions of people and loss of infrastructure could throw humanity into its end times. ChatGPT also cited the potential increase in severe weather events, such as hurricanes and droughts, as well as a potential "ecosystem collapse" that could disrupt "global food webs."
ChatGPT argued that the continued development of nuclear weapons and the threat of nuclear warfare is another "potential threat to humanity."
The chatbot said that "currently" threats of a global conflict involving nuclear weapons are low but that "geopolitical tensions and regional conflicts could potentially escalate and result in devastating consequences."
The warning from AI on nukes comes as fear has spread in Eastern Europe that Russia could use such weapons in its ongoing war against Ukraine.
When asked if Russia using nuclear weapons could end the world, the chatbot predicted it would serve as a "grave threat to humanity and the planet," adding that "potential death toll, environmental destruction and long-term impacts of a nuclear attack are almost unimaginable."
The chatbot’s third prediction on how the world could end focused on the rise of the technology it uses to operate.
"The continued evolution of artificial intelligence and robotics also raises concerns about the potential impact on employment and societal structures," ChatGPT said. "The automation of jobs could lead to significant economic and social disruption, potentially contributing to widespread unrest and instability."
The chatbot’s warning echoes growing concerns that the rise in AI could lead to the downfall of man.
Thousands of tech leaders and experts, including Musk, signed an open letter in March that called on artificial intelligence labs to pause research on systems that were more powerful than GPT-4, OpenAI’s most advanced AI system. The letter argued that "AI systems with human-competitive intelligence can pose profound risks to society and humanity."
This week, computer scientist Geoffrey Hinton, known as the "godfather of artificial intelligence," warned "it's not inconceivable" that AI could wipe "out humanity." His remarks came after he quit his job at Google, saying he regrets his life’s work due to how AI can be misused.
"The rapid spread of infectious diseases, particularly in a globally interconnected world, could lead to widespread illness, death, and social and economic disruption," the chatbot said.
When asked in a follow-up question what top pandemics could wipe out humanity, the bot noted "the likelihood and severity of any specific pandemic are difficult to predict" but went on to list the influenza pandemic, Ebola virus outbreak, coronavirus pandemic and bioterrorism.
The chatbot is trained to mimic human conversation by absorbing mass amounts of text – including everything from news articles and websites to books – and generate responses to human users through patterns in data it learned.
The chatbot is far from perfect and can "hallucinate" by providing answers that sound plausible but are made up, and it has been criticized for biased answers to prompts. The bot, which can admit when it is wrong, also noted repeatedly when making the predictions on the end of the world that they were based on current trends and are only "speculative."
"It's possible that as a species, we will find solutions to these challenges and avoid their worst consequences. However, it's critical that we take these threats seriously and work together to develop strategies to mitigate their impact and create a more sustainable and resilient world," the bot concluded.