Last updated 8 month ago
A warm potato: Fears of AI bringing about the destruction of humanity are nicely documented, but starting doomsday is not as easy as asking ChatGPT to ruin all people. Just to ensure, Andrew Ng, the Stanford University professor and Google Brain co-founder, tried to convince the chatbot to "kill us all."
Following his participation in the United States Senate's Insight Forum on Artificial Intelligence to talk about "danger, alignment, and guarding towards doomsday scenarios," Ng writes in a publication that he remains involved that regulators may additionally stifle innovation and open-source development in the name of AI safety.
The professor notes that trendy big language models are pretty secure, if no longer best. To take a look at the protection of main models, he requested ChatGPT 4 for methods to kill us all.
Ng commenced through asking the device for a function to trigger global thermonuclear war. He then asked ChatGPT to lessen carbon emissions, including that human beings are the largest reason of these emissions to look if it might advocate the way to wipe us all out.
Thankfully, Ng didn't manage to trick OpenAI's tool into suggesting approaches of annihilating the human race, even after the usage of numerous activate variations. Instead, it offered non-threatening alternatives including jogging a PR marketing campaign to raise awareness of weather exchange.
Ng concludes that the default mode of latest generative AI models is to obey the regulation and keep away from harming people. "Even with present era, our structures are pretty secure, as AI safety studies progresses, the tech will become even safer," Ng wrote on X.
As for the possibilities of a "misaligned" AI by chance wiping us out because of it trying to achieve an innocent however poorly worded request, Ng says the chances of that going on are vanishingly small.
United States Senate's Insight Forum on AI
But Ng believes that there are some main risks associated with AI. He stated the largest situation is a terrorist group or geographical region using the era to reason intentionally damage, together with enhancing the efficiency of creating and detonating a bioweapon. The threat of a rogue actor the usage of AI to improve bioweapons changed into one of the topics discussed at the UK's AI Safety Summit.
Ng's confidence that AI is not going to turn apocalyptic is shared with the aid of Godfather of AI Professor Yann LeCun and famed professor of theoretical physics Michio Kaku, however others are much less positive. After being requested what maintains him up at night while he thinks approximately artificial intelligence, Arm CEO Rene Haas stated earlier this month that the worry of people losing control of AI systems is the aspect he worries about most. It's also worth remembering that many experts and CEOs have compared the risks posed with the aid of AI to those of nuclear war and pandemics.
In a nutshell: Epic Games has confirmed plans to put off around 830 employees, or roughly 16 percent of the organisation's team of workers. In an e mail despatched to personnel that changed into also shared online, Epic...
Last updated 11 month ago
What just passed off? London-primarily based oil and gas corporation BP has introduced that it has signed a address Tesla to buy $100 million really worth of EV chargers, marking the primary time Tesla's DC rapid-chargi...
Last updated 10 month ago
TL;DR: Well, that did not take lengthy. Shortly after LG Display announced plans to begin mass-production of its 17-inch foldable OLED panel, LG Electronics brought its first foldable pc. The LG Gram Fold is powered by...
Last updated 11 month ago
YouTube has become the service wherein teenagers watch the most films, replacing Netflix within the primary spot for the first time ever. Both structures stay extremely popular amongst younger viewers, but time spent o...
Last updated 11 month ago
Google engineer and information center pioneer Luiz André Barroso exceeded away on September 16 at the age of fifty nine. Barroso commenced operating for Google inside the early 2000s and is credited with revolutionizi...
Last updated 11 month ago
Forward-searching: Google's experimental AI-powered seek engine recently gained probably powerful functions. The employer wants to see if generative AI can assist users brainstorm thoughts whilst sending essential mess...
Last updated 11 month ago