AI & Robotics News and Discussions

User avatar
caltrek
Posts: 6509
Joined: Mon May 17, 2021 1:17 pm

Re: AI & Robotics News and Discussions

Post by caltrek »

How Does ChatGPD Differ From Human Intelligence?
by Corrie Pikul-Browm
February 15, 2023

Extract:
(Futurity) the backbone of ChatGPT is a state-of-the-art kind of artificial neural network called a transformer network. These networks, which came out of the study of natural language processing, have recently come to dominate the entire field of artificial intelligence. Transformer networks have a particular mechanism that computer scientists call “self-attention,” which is related to the attentional mechanisms that are known to take place in the human brain.

Another similarity to the human brain is a key aspect of what has enabled the technology to become so advanced, (Thomas) Serre (professor of cognitive, linguistic, and psychological sciences and of computer science) said. In the past, he explained, training a computer’s artificial neural networks to learn and use language or perform image recognition would require scientists to perform tedious, time-consuming manual tasks like building databases and labeling categories of objects.

Modern large language models, such as the ones used in ChatGPT, are trained without the need for this explicit human supervision. And that seems to be related to what Serre referred to as an influential brain theory known as the predictive coding theory. This is the assumption that when a human hears someone speak, the brain is constantly making predictions and developing expectations about what will be said next.

On a darker note, in the same way that the human learning process is susceptible to bias or corruption, so are artificial intelligence models. These systems learn by statistical association, Serre said. Whatever is dominant in the data set will take over and push out other information.

NOPE, CHATGPT DOESN’T DREAM

One area in which human brains and neural networks diverge is in sleep—specifically, while dreaming. Despite AI-generated text or images that seem surreal, abstract, or nonsensical, (Elle) Pavlick (assistant professor of computer science and a research scientist at Google AI) said there’s no evidence to support the notion of functional parallels between the biological dreaming process and the computational process of generative AI. She said that it’s important to understand that applications like ChatGPT are steady-state systems—in other words, they aren’t evolving and changing online, in real-time, even though they may be constantly refined offline.
Read more here: https://www.futurity.org/author/corrie-pikul/


For more discussion on how artificial intelligence models may be susceptible to bias or corruption: https://www.futurity.org/stroke-risk-a ... 863922-2/
Don't mourn, organize.

-Joe Hill
weatheriscool
Posts: 12967
Joined: Sun May 16, 2021 6:16 pm

Re: AI & Robotics News and Discussions

Post by weatheriscool »

ChatGPT able to pass Theory of Mind Test at 9-year-old human level

by Bob Yirka , Tech Xplore
https://techxplore.com/news/2023-02-cha ... human.html
Michal Kosinski, computational psychologist at Stanford University, has been testing several iterations of the ChatGPT AI chatbot developed by Open AI on its ability to pass the famous Theory of Mind Test. In his paper posted on the arXiv preprint server, Kosinski reports that testing the latest version of ChatGPT found that it passed at the level of the average 9-year-old child.

ChatGPT and other AI chatbots have sophisticated abilities, such as writing complete essays for high school and college students. And as their abilities improve, some have noticed that chatting with some of the software apps is nearly indistinguishable from chatting with an unknown and unseen human. Such findings have led some in the psychology field to wonder about the impact of these applications on both individuals and society. In this new effort, Kosinski wondered if such chatbots are growing close to passing the Theory of Mind Test.

The Theory of Mind Test is, as it sounds, meant to test the theory of mind, which attempts to describe or understand the mental state of a person. Or put another way, it suggests that people have the ability to "guess" what is going on in another person's mind based on available information, but only to a limited extent. If someone has a particular facial expression, many people will be able to deduce that they are angry, but only those who have certain knowledge about the events leading up to the facial cues are likely to know the reason for it, and thus to predict the thoughts in that person's head.
User avatar
funkervogt
Posts: 1171
Joined: Mon May 17, 2021 3:03 pm

Re: AI & Robotics News and Discussions

Post by funkervogt »

This essay about the future of AI is too alarmist, IMO, but it makes two good points that I agree with:

1) Thanks to recent developments like ChatGPT, expert forecasts about what year AGI will be invented have either stayed the same or shrunk. No one's forecast has gotten longer.
2) Our ability to keep AI aligned with human goals is poor and will likely remain so. Few computer experts are working on the problem, and there have been countless examples of humans finding ways to cleverly mislead the existing narrow AIs to get them to circumvent their existing alignment safeguards (e.g. - trick it into drawing a nude woman or to make a racist joke).

https://musingsandroughdrafts.com/2023/ ... f-ai-risk/
firestar464
Posts: 672
Joined: Wed Oct 12, 2022 7:45 am

Re: AI & Robotics News and Discussions

Post by firestar464 »

https://apnews.com/article/technology-m ... e32d8b91eb

"Fixed?" Just wait till some nerd in their basement finds some prompt to unleash Riley and her homies Jack, Nelson, and Sade. /j

Seriously though, people will continue to find new exploits and new secrets, and this will be interesting for further Bing and AI research.
weatheriscool
Posts: 12967
Joined: Sun May 16, 2021 6:16 pm

Re: AI & Robotics News and Discussions

Post by weatheriscool »

Zuckerberg Introduces Meta's Answer to ChatGPT, LLaMA
Source: Gizmodo
Move over OpenAI’s ChatGPT, Google’s Bard, and Microsoft’s Prometheus—there’s yet another large language model-powered artificial intelligence in town. Meta, the company formerly known as Facebook, introduced its own AI today, called LLaMA.

-snip-

“Today we’re releasing a new state-of-the-art AI large language model called LLaMA designed to help researchers advance their work,” Zuckerberg wrote on his social media platform. He added that “LLMs have shown a lot of promise in generating text, having conversations, summarizing written material, and more complicated tasks like solving math theorems or predicting protein structures.” But the Meta exec did not explain exactly which (if any) of those tasks LLaMA could currently accomplish.

In fact, the only detail that Zuckerberg offered on the large language model in the Friday announcement is that his company is “committed to this open model of research and we’ll make our new model available to the AI research community.”

Gizmodo reached out to Meta with questions about LLaMA’s capacity, integration into the company’s products, and if/when/how the AI will be made publicly available, but did not immediately receive a response.
-snip-

Read more: https://gizmodo.com/facebook-chatgpt-go ... 1850155514
weatheriscool
Posts: 12967
Joined: Sun May 16, 2021 6:16 pm

Re: AI & Robotics News and Discussions

Post by weatheriscool »

Microsoft researchers are using ChatGPT to instruct robots and drones (with video)
OpenAI's ChatGPT isn't just good at generating coherent text responses to natural language prompts -- it can also play a role in human-to-robot interactions and use sensor feedback to write code for robot actions.
Microsoft recently conducted research to "see if ChatGPT can think beyond text, and reason about the physical world to help with robotics tasks." The aim was to see if people can use ChatGPT to instruct robots without learning programming languages or understanding robotic systems.


"The key challenge here is teaching ChatGPT how to solve problems considering the laws of physics, the context of the operating environment, and how the robot's physical actions can change the state of the world," a team from Microsoft Autonomous Systems and Robotics Research note in a blogpost.


The researchers note in their blog post: "ChatGPT asked clarification questions when the user's instructions were ambiguous, and wrote complex code structures for the drone such as a zig-zag pattern to visually inspect shelves."

Microsoft tested ChatGPT to use a robotic arm to move blocks around to form the Microsoft logo. The researchers also tasked ChatGPT with writing an algorithm for a drone to reach a point without crashing into obstacles. They also tested whether ChatGPT can decide where a robot should go based on sensor feedback in real time.
https://www.zdnet.com/article/microsoft ... nd-drones/
firestar464
Posts: 672
Joined: Wed Oct 12, 2022 7:45 am

Re: AI & Robotics News and Discussions

Post by firestar464 »

SkynetGPT
spryfusion
Posts: 394
Joined: Thu Aug 19, 2021 4:29 am

Re: AI & Robotics News and Discussions

Post by spryfusion »

User avatar
wjfox
Site Admin
Posts: 8732
Joined: Sat May 15, 2021 6:09 pm
Location: London, UK
Contact:

Re: AI & Robotics News and Discussions

Post by wjfox »

Artificial intelligence is on the brink of an 'iPhone moment' and can boost the world economy by $15.7 trillion in 7 years, Bank of America says

Phil Rosen
Mar 1, 2023, 1:50 PM

Artificial intelligence is about to have its "iPhone moment" and could revolutionize everything, according to Bank of America.

In a Tuesday note to clients, BofA strategists listed four reasons why AI is about to change the landscape: democratization of data, unprecedented mass adoption, "warp-speed" technological development, and abundant commercial uses.

"We are at a defining moment - like the internet in the '90s - where Artificial Intelligence (AI) is moving towards mass adoption, with large language models like ChatGPT finally enabling us to fully capitalize on the data revolution," they said.

Up until recently, AI could read and write but couldn't understand content, BofA said. Tools like ChatGPT have changed that, however, and its ability to understand natural language has opened the door to huge upside.

[...]

"Our estimate may prove to be conservative as growth in [large language models] and other generative AI technologies could be even faster than we expect given advancements in machine learning and deep learning capabilities," UBS Global Wealth Management's Solita Marcelli wrote in a client note Tuesday.

https://markets.businessinsider.com/new ... ets-2023-3
firestar464
Posts: 672
Joined: Wed Oct 12, 2022 7:45 am

Re: AI & Robotics News and Discussions

Post by firestar464 »

"You are not a parrot, and a chatbot is not a human."

https://nymag.com/intelligencer/article ... ender.html
Post Reply