Business

The true purpose of AI may not be intelligence

The true purpose of AI may not be intelligence

The true purpose of AI may not be intelligence

The true purpose of AI may not be intelligence

AI is rapidly finding industrial applications, such as using large language models to automate enterprise IT. Such applications may make the question of true intelligence moot.

Tiernan Ray/ZDNET

British mathematician Alan Turing he wrote In 1950, “Can machines think?’ I suggest you ponder the question.His research marked the decades-long debate in artificial intelligence research.

For a couple of generations of scientists contemplating AI, the question of whether “true” or “human” intelligence could be achieved was always an important part of the work.

AI may now be at a tipping point where questions like these matter less and less to most people.

The emergence of something called industrial AI in recent years may signal the end of these major concerns. AI is more capable today More than 66 years since computer scientist John McCarthy first coined the term AI. As a result, the industrialization of AI is shifting the focus from intelligence to achievement.

As well: OpenAI’s Dall•E 2 means we never need stock photos again

These achievements are remarkable. They include a system that can predict protein folding, AlphaFoldFrom Google’s DeepMind unit and text generator GPT-3 From the launch of OpenAI. Both of those programs have tremendous industrial promise, whether or not anyone calls them smart.

Among other things, AlphaFold holds the promise of designing new protein shapes, a possibility that has electrified the biology community. GPT-3 is quickly gaining popularity as a system that can automate business tasks, such as answering written questions from employees or customers without human intervention.

This practical success, fueled by a thriving field of semiconductors, Led by chipmaker Nvidiahe seems to be able to overcome his old preoccupation with intelligence.

In one corner of industrial AI, no one cares whether such programs will achieve intelligence. it’s like it’s in front of you practical achievements that shows obvious value, the old question, “But is it smart?” ceases to matter.

As well: AI Critic Gary Marcus: Meta’s LeCun finally comes back to things I said years ago

Hector Levesque as computer scientist has writtenthen science Versus AI technology“Unfortunately, it’s AI technology that gets all the attention.”

To be sure, the question of true intelligence still matters to a few thinkers. Over the past month, ZDNET has interviewed two prominent scholars who are deeply concerned with that question.

Yann LeCun, chief AI scientist at Facebook owner Meta Properties, he spoke at length He talked to ZDNET about a paper he released this summer as a think piece on where AI should go. LeCun expressed concern about the main work of deep learning today, if it continues on its current path, he will not attain what he calls “true” intelligenceincluding the ability of a computer system to plan a course of action using common sense.

LeCun expresses the concern of an engineer without real intelligence that these programs will ultimately be brittle, meaning they can break before they do what we want them to do.

As well: Meta’s AI guru LeCun: Most current AI approaches will never lead to true intelligence

“You know, I think it’s entirely possible to have Level 5 autonomous cars without meaning to,” LeCun told ZDNET, referring to efforts by Waymo and others to build ADAS (Advanced driver assistance systems) for self-driving, “but you’ll have to engineer the hell out of it.”

And NYU professor emeritus Gary Marcus, a frequent critic of deep learning, he told ZDNET this month That it is stuck as an AI field to find anything resembling human intelligence.

“I don’t want to question whether it’s intelligence or not,” Marcus told ZDNET. “But the form of intelligence that we might call general intelligence or adaptive intelligence, I care about adaptive intelligence. […] We don’t have such a machine.”

Heads of Yann LeCun and Gary Marcus

Meta team’s Yann LeCun (right) and AI critic Gary Marcus.

Both LeCun’s and Marcus’s concerns seem more and more outlandish. Industrial AI professionals don’t want to ask hard questions, they just want things to work well. As more and more people put AI in their hands, data scientists and self-driving car engineers, people removed from the basic scientific questions of research, can ask, “Can machines think?” becomes more important.

Even scientists who are aware of AI’s shortcomings are tempted to ignore it in order to enjoy the practical utility of the technology.

As well: The best travel agent is an AI algorithm

More scholarly than Marcus or LeCun, but mindful of the dichotomy of practical and profound, is Demis Hassabis, co-founder of DeepMind.

At conference in 2019 At the Institute for Advanced Study in Princeton, New Jersey, Hassabis pointed out the limitations of many AI programs that could only do one thing well, like an idiot scholar. DeepMind, Hassabis said, is trying to develop a broader and richer capability. “We are trying to find a meta-solution to solve other problems,” he said.

However, Hassabis loves DeepMind’s latest inventions just as much as the particular tasks that highlight them.

When DeepMind recently introduced an improved way do linear algebramathematics at the heart of deep learning, Hassabis praised the achievement to the exclusion of any pretensions to intelligence.

“It turns out that everything is a matrix multiplication, from computer graphics to training neural networks,” Hassabis. he wrote on Twitter. Maybe that’s true, but it allows us to abandon the search for intelligence in favor of improving a tool, as if to say: If it works, why ask why?

The field of AI is undergoing a shift in attitude. There was a time when every achievement of an AI program, no matter how good, was greeted with the skeptical remark: “Well, that doesn’t mean it’s smart.” It’s a model that AI historian Pamela McCorduck has called.moving the doors.”

Today, things seem to be the other way around: people tend to casually assign intelligence to anything labeled AI. If a chatbot like Google’s LAMDA generates enough natural language sentences, someone will argue that it is sensitive.

Alan Turing 16 years

British mathematician Alan Turing predicted that “educated general opinion” would come to accept that machines have intelligence.

Turing himself anticipated this change of attitude. He predicted that ways of talking about computers and intelligence would change in favor of computer acceptance behaviour as intelligent

“I believe that by the end of the century the usage of words and educated general opinion will have changed so much that machines will be able to speak what they think without expecting contradictions,” wrote Turing.

As the honest question of intelligence fades, the mere rhetoric of intelligence is allowed to float freely in society for other agendas.

As well: Nvidia CEO Jensen Huang: AI language models as a service ‘may be one of the biggest software opportunities ever’

In a mixed brilliant encomio in Fast Company recentlyWritten by Michael Hochberg, a computer industry executive, and Robert Spalding, a retired Air Force general, the authors make modest assertions about intelligence as a way to add organ music to a dire warning of geopolitical danger:

The stakes couldn’t be higher when training general artificial intelligence systems. AI is the first tool that convincingly replicates the unique capabilities of the human mind. It has the ability to create a unique and targeted user experience for each citizen. This may be the ultimate propaganda tool, a weapon of deception and persuasion that has never existed in history.

Most scholars would agree.”artificial general intelligence,” while it makes sense as a term, it’s nowhere close to achieving today’s technology. Hochberg and Spalding’s claims about what the program can do are exaggerated.

Honest assertions about what AI is doing hide the nuances of individuals like LeCun and Marcus. A rhetorical regime is emerging that is concerned with persuasion, not intelligence.

That may be the direction of things for the foreseeable future. If AI does more and more things in biology, physics, business, logistics, marketing and warfare, and as society becomes comfortable with it, fewer and fewer people may even care to ask. But is it smart?



#true #purpose #intelligence

مقالات ذات صلة

اترك تعليقاً

لن يتم نشر عنوان بريدك الإلكتروني. الحقول الإلزامية مشار إليها بـ *

زر الذهاب إلى الأعلى
سيتات آورج 2022 سيتات آورج 2022