{"id":14761,"date":"2024-04-10T10:10:09","date_gmt":"2024-04-10T10:10:09","guid":{"rendered":"http:\/\/TheNextWeb=1405572"},"modified":"2024-04-10T10:10:09","modified_gmt":"2024-04-10T10:10:09","slug":"metas-ai-chief-llms-will-never-reach-human-level-intelligence","status":"publish","type":"post","link":"https:\/\/www.londonchiropracter.com\/?p=14761","title":{"rendered":"Meta\u2019s AI chief: LLMs will never reach human-level intelligence"},"content":{"rendered":"\n

There\u2019s no escaping the hype around artificial general intelligence. Barely a day passes without a new headline<\/a> about the concept, which envisions computer systems outperforming humans at various cognitive tasks.<\/p>\n

In the last month alone, a trio of tech luminaries have added fresh proclamations. Nvidia CEO Jensen Huang suggested<\/a> AGI would arrive within five years. Ben \u201cfather of AGI\u201d Goertzel <\/span>forecasted<\/span><\/a> a mere three. <\/span>Elon Musk <\/span>typically<\/span><\/a> made the boldest prediction for the tipping point: the end of 2025<\/span>.<\/span><\/p>\n

Still, not everyone is so bullish. One notable sceptic is Yann LeCun, Meta\u2019s chief AI scientist and a winner of the prestigious Turing Award.<\/span><\/p>\n

Often referred to as one of three \u201cgodfathers of AI,\u201d LeCun goes as far as to argue<\/span> that <\/span>\u201cthere is no such thing as AGI\u201d because \u201chuman intelligence is nowhere near general.\u201d The Frenchman prefers<\/a> to chart a path towards \u201chuman-level AI.\u201d <\/span><\/span><\/p>\n

\n
\n
<\/div>\n

\n

The <3 of EU tech<\/p>\n

The latest rumblings from the EU tech scene, a story from our wise ol’ founder Boris, and some questionable AI art. It’s free, every week, in your inbox. Sign up now!<\/p>\n<\/div>\n<\/div>\n<\/div>\n

At an event on Tuesday in London \u2014 Meta\u2019s flagship engineering hub outside the US \u2014 LeCun said even that remains a distant destination.<\/p>\n

He pointed to a quartet of cognitive challenges: reasoning, planning, persistent memory, and understanding the physical world.<\/p>\n

\u201cThose are four essential characteristics of human intelligence \u2014 also animal intelligence, for that matter \u2014 that current AI systems can\u2019t do,\u201d he said.<\/p>\n

Without these capabilities, AI<\/a> applications remain limited and error-prone. Autonomous vehicles still aren\u2019t safe for public roads. Domestic robots struggle with basic household chores. Our smart assistants can only complete basic tasks.<\/p>\n

These intellectual shortcomings are particularly prominent in large language models (LLMs). In LeCun\u2019s view, they\u2019re severely restricted by their reliance on one form of human knowledge: text.<\/p>\n

\u201cWe\u2019re easily fooled into thinking they are intelligent because of their fluency with language, but really, their understanding of reality is very superficial,\u201d he said.<\/p>\n

\u201cThey\u2019re useful, there\u2019s no question about that. But on the path towards human-level intelligence, an LLM is basically an off-ramp, a distraction, a dead end.\u201d<\/p>\n

Why LLMs aren\u2019t as smart as they seem<\/h2>\n

The likes of Meta\u2019s LLaMA, OpenAI\u2019s<\/a> GPT-3, and Google\u2019s Bard are trained on enormous quantities of data. According to LeCun, it would take a human around 100,000 years to read all the text ingested by a leading LLM. But that\u2019s not our primary method of learning.<\/p>\n

We consume far more information through our interactions with the world. LeCun estimates that a typical four-year-old has seen 50 times more data<\/a> than the world\u2019s biggest LLMs.<\/p>\n

\u201cMost of human knowledge is actually not language so those systems can never reach human-level intelligence \u2014 unless you change the architecture,\u201d LeCun said.<\/p>\n

Naturally, the 63-year-old has an alternative architecture. He calls it \u201cobjective-driven AI.\u201d<\/p>\n

Objectives of intelligence<\/h2>\n

Objective-driven AI systems are built to fulfil specific goals set by humans.<\/p>\n

Rather than being raised on a diet of pure text, they learn about the physical world through sensors and training on video data.<\/p>\n

The result is a \u201cworld model\u201d that shows the impact of actions. All the potential changes are then updated in the system\u2019s memory.<\/p>\n

What would be the difference, for instance, if a chair is pushed to the left or the right of a room? By learning through experience, the end states begin to become predictable. As a result, machines can plan the steps needed to complete various tasks.<\/p>\n

LeCun is quietly confident about the pay-off.<\/span><\/p>\n

\u201cEventually, machines will surpass human intelligence\u2026 it\u2019s gonna take a while though,\u201d he said. \u201cIt\u2019s not just around the corner \u2014 and it\u2019s certainly not next year like our friend Elon has said.\u201d<\/p>\n

One of the themes of this year\u2019s TNW Conference is Ren-AI-ssance: The AI-Powered Rebirth. If you want to go deeper into all things artificial intelligence, or simply experience the event (and say hi to our editorial team), we\u2019ve got something special for our loyal readers. Use the code TNWXMEDIA at checkout to get 30% off your <\/span><\/i>business pass<\/span><\/i><\/a>, <\/span><\/i>investor pass<\/span><\/i><\/a> or startup packages (<\/span><\/i>Bootstrap<\/span><\/i><\/a> & <\/span><\/i>Scaleup<\/span><\/i><\/a>).<\/span><\/i><\/p>\n

Source<\/a><\/p>\n","protected":false},"excerpt":{"rendered":"

There\u2019s no escaping the hype around artificial general intelligence. Barely a day passes without a new headline about the concept, which envisions computer systems outperforming humans at various cognitive tasks. In the last…<\/p>\n","protected":false},"author":1,"featured_media":14762,"comment_status":"open","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":[],"categories":[1],"tags":[],"_links":{"self":[{"href":"https:\/\/www.londonchiropracter.com\/index.php?rest_route=\/wp\/v2\/posts\/14761"}],"collection":[{"href":"https:\/\/www.londonchiropracter.com\/index.php?rest_route=\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.londonchiropracter.com\/index.php?rest_route=\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.londonchiropracter.com\/index.php?rest_route=\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/www.londonchiropracter.com\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=14761"}],"version-history":[{"count":0,"href":"https:\/\/www.londonchiropracter.com\/index.php?rest_route=\/wp\/v2\/posts\/14761\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.londonchiropracter.com\/index.php?rest_route=\/wp\/v2\/media\/14762"}],"wp:attachment":[{"href":"https:\/\/www.londonchiropracter.com\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=14761"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.londonchiropracter.com\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=14761"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.londonchiropracter.com\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=14761"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}