The pace of development and release of new artificial intelligence models appears to be accelerating as more applications appear every day. What was once limited to an obscure pocket of technology researchers has become mainstream with major companies, like Microsoft and OpenAI, investing heavily in the future of AI.
Microsoft Vice President Ashley Llorens says current AI models can now handle very difficult tasks almost automatically. Reasoning and information processing abilities have reached unprecedented levels.
“What we have now is we have AI that can reason better, that can perceive the environment in a more sophisticated way,” Llorens said. Yahoo! Finance. “And so, that means we will be able to delegate a more sophisticated set of tasks to AI to complete on our behalf.”
Multimodal models, capable of interpreting and creating results using text, images or videos, will be more prevalent in the near future. An example includes Microsoft’s co-pilot visiona Copilot feature for Windows. Among other things, the application would be able to understand a web page and answer all users’ questions about it.
Currently, artificial intelligence data centers consume a lot of energy to operate. Llorens predicts that this will change through innovation and hardware improvements, with servers eventually becoming more efficient and more environmentally friendly.
AI will be “used for everything”
Sam Altman, founder and CEO of OpenAI, also has predictions about artificial intelligence. Altman shares Llorens’ view that the technology will be able to accomplish very difficult tasks and will likely use a variety of other tools to accomplish them.
At the New York Times DealBook Summit in New York, Altman said advances in AI “superintelligence” won’t have much impact at first. Yet ultimately, it “will be more intense than people think,” implying that there could be significant employment disruption ahead.
“There will be incredibly good (AI) models, widely available, used for everything,” he said.
When asked about the potential dangers of OpenAI’s ChatGPT, Altman defended the chatbot. For now, ChatGPT is “secure enough and robust enough” by most users’ standards, Altman noted. However, many people fear unbridled growth in AI technology will be a risk to humanity and encourage government intervention before it is too late.
Altman seems to agree. If ever AI exceeds human intelligence“We are going to have to have a certain confidence in our governments,” he stressed. Altman confidently assumes that the world’s governments will communicate and address any challenges created by AI.