Wednesday, June 12, 2024
No menu items!
HomeGoogle NewsHas GPT-4 Really Passed the Startling Threshold of Human-Level Artificial Intelligence? Well,...

Has GPT-4 Really Passed the Startling Threshold of Human-Level Artificial Intelligence? Well, It Depends

The recent public interest in tools like ChatGPT has raised an old question in the artificial intelligence community: whether artificial general intelligence (in this case, AI works at the human level) person) can it be achieved? An online preprint this week has added to the hype, showing that the latest cutting-edge large language model, GPT-4, is in the early stages of generalized artificial intelligence (AGI) as it represents “the rays of intelligence”.

OpenAI, the company behind ChatGPT, did not hesitate to announce its pursuit of AGI. Meanwhile, a large number of researchers and public intellectuals have called for an immediate halt to the development of these models, citing “profound risks to society and humanity”.

These calls to halt AI research are just theatrical and unlikely to succeed – the allure of advanced intelligence is too provocative for humans to ignore and too rewarding for companies to temporarily shut down. stop.

  • Microsoft puts ads on AI-powered Bing Chat Search Engine

But are the worries and hopes about AGI warranted? How close are GPT-4 and broader AI to general human intelligence? If human cognitive abilities are a landscape, AI has indeed increasingly taken over large swaths of this territory.

It can now perform many distinct cognitive tasks better than humans in the areas of vision, image recognition, reasoning, reading comprehension, and game play.

These AI skills have the potential to lead to a realignment of the global labor market in less than ten years.

But there are at least two ways of looking at the AGI problem.

The first is that over time, AI will develop human-appropriate skills and learning abilities and reach AGI levels.

The expectation is that the unique human ability to continuously grow, learn, and transfer learning from one domain to another will eventually be mirrored by AI.

  • Elon Musk, Experts warn against advanced AI, cites risks to humanity

This is in contrast to current AI, where training in one area, such as detecting cancer in medical imaging, does not transfer to other areas.

So the concern of many is that at some point, AI will surpass human intelligence, and then quickly overshadow us, making us appear before the AIs of the future. like the ants appearing before our eyes now.

The plausibility of AGI is disputed by some philosophers and researchers, citing that current models are largely ignorant of the outputs (that is, they do not understand what they are producing).

They also have no prospect of gaining consciousness because they are primarily predictive – automating what will happen next in text or other outputs.

Instead of becoming intelligent, these models simply recombine and replicate the data they were trained on.

Consciousness, the essence of life, is missing. Even as AI-based models continue to evolve and accomplish more sophisticated tasks, there is no guarantee that consciousness or AGI will emerge. And if it does appear, how do we recognize it? ChatGPT’s usefulness and GPT-4’s ability to master some tasks equal to or better than humans (such as bar exams and academic olympiads) give the impression that AGI is very close.

See also  Onida 1.5 Ton 3 Star Inverter Split AC (IR183PRS)

This view is confirmed by the rapid performance improvement with each new model.

Now, there is no doubt that AI can outperform humans in many individual cognitive tasks. There is growing evidence that the best model for interacting with AI may be a human/machine pairing model – where our own intelligence is augmented rather than superseded by AI.

Signs of such pairing have appeared with announcements about the work of co-pilots and programmers pairing AI to code.

It seems almost inevitable that the future of our work, life, and school will have the pervasive and persistent presence of AI.

According to that metric, the possibility of AI being considered intelligent is reasonable, but this is still a controversial space and many people have opposed it. The famous linguist Noam Chomsky has stated that the day of AGI “may come, but the dawn has not yet come”.

The second angle is to look at the idea of ​​intelligence as it is practiced by people in their daily lives. According to one school of thought, we are intelligent primarily in networks and systems rather than as solitary individuals. We hold knowledge in networks.

Until now, those networks were mostly human. We can get insight from someone (like the author of a book), but we don’t see them as an active “agent” in our perception.

But ChatGPT, Copilot, Bard, and other AI-powered tools can become part of our cognitive network – we interact with them, ask them questions, they restructure documents and resources for us. In this sense, AI does not need to have sentience or general intelligence.

See also  TCL 50 XL Nxtpaper 5G

It simply needs the ability to be embedded in and part of our knowledge network to replace and complement many of our current jobs and tasks.

The current focus on AGI has overlooked many of the opportunities that current models and tools present to us. Sentimentally, consciously or not – all these attributes are irrelevant to the many people who are using AI to co-create art, organize articles and essays, develop videos, and navigate their lives. living.

The most relevant or most pressing concern for humans is not whether AI is intelligent when it comes to working on its own and not connecting to humans.

It can be argued that today we are smarter, more competent and more creative with AI as it enhances our cognitive capacities. Right now, it seems like the future of humanity could be in partnership with AI – a journey well underway.

From smartphones with rollable or liquid-cooled displays, to compact AR glasses and handheld devices that can be easily repaired by their owners, we discuss the best devices The best we’ve seen at MWC 2023 on Orbital, the Gadgets 360 podcast. Orbital is available on Spotify, Gaana, JioSaavn, Google Podcasts, Apple Podcasts, Amazon Music, and wherever you download your podcasts.
Affiliate links may be automatically generated – see our ethics statement for details.


Please enter your comment!
Please enter your name here

Most Popular

HTC U24 Pro

HMD 110

HMD 105

Acer ALG Gaming Laptop

Realme GT 6

Huawei MatePad SE 11

Xiaomi 14 Civi