ArticlesArtificial Intelligence

AI safety: Do reliability factors that tamed animals apply to LLMs?

3 Mins read
LLMs

Humans know that animals can’t speak advanced languages, even though they find ways to communicate. Humans are also ensured that the skills of animals to notice or analyze things around them are limited. They might repeat what they see, which is a basic way to learn, but it does not offer much to those within or outside their habitat. They can learn new things if we train them, but only up to a point.

Animals are strong in their own ways, yet they were tamed by humans. Humans did this mostly through advanced language and intelligence. Motor skills were also important with bipedalism, however, birds have something like this, too.

AI, or artificial intelligence, is digital. It doesn’t make choices on its own—it has to be told what to do. You can turn it off.  It does not have real world experiences like we do. It cannot feel, know, or think. It does not have understanding. It does not have intelligence. It is also not sentient. It’s not aware or conscious. Some people say it’s just a machine that does what it’s told.

Even so, AI can get better at what it does if it uses the smart things it has, not just the smart things it knows about. Books are full of smart ideas, but they can’t use those ideas themselves. Many animals hear people talk all the time. They live in the same world we do, but they can’t use all that they learn to make their lives or the lives of other animals better.

AI is in digital. It has learned a lot about what people know and can do. It can do what we expect, and sometimes it can do even more. It can learn from text, images, audio, and videos, and then it can change how it acts or give advice.

Computers can show us the world almost as well as we can see it through a window. For AI, computers are like windows that help it understand our world.

Our minds treat everything the same, whether it’s something we see on a screen or right in front of us. Computers are a big part of what we do for work and fun. AI lives in the computer world. It’s getting better, learning new things, and it can use words like we do.

AI can do a lot to help us work better and faster. And because so much of what humans do with friends is on computers now, it can help us talk to each other, too. Humans rely on AI so much that it would be hard to stop using it, even if they wanted to.

If animals could be smarter, or if humans could make things that let them use their intelligence, they would probably do better. But this might not be true for AI, especially the kind that works with language. Computers are so important nowadays that AI might become even more common. The dynamism of LLMs might define it within digital, which includes new types of robots. Humans can train AI, and it is not just the guesswork that remains with animals.

Could Animal Intelligence and LLMs be Compared?

What does an animal think when it sees a jetliner or a PC? If it’s something an animal can’t touch or smell, does it understand it? Though animals have great intelligence, experientially, but they have blanks a lot of times. They usually can’t learn things that aren’t right in front of them.

LLMs doesn’t have the experience of the external world. But it has a lot of information about it. This doesn’t mean LLMs are directly comparable with animals. But AI might be able to do things that living creatures can’t in terms of extending capabilities.

David Stephen does research in conceptual brain science, centered on the electrical and chemical impulses of nerve cells, for a model of the human mind, applicable to mental health, consciousness, neurotechnology and artificial intelligence. He was a visiting scholar in medical entomology at the University of Illinois Urbana-Champaign, UIUC, IL.

Read Next: Massive AI opportunity for 2 million Indians as Microsoft announces ADVANTA(I)GE INDIA initiative

Leave a Reply

Your email address will not be published. Required fields are marked *

÷ 6 = 1