Artificial General Intelligence (AGI) News and Discussions

[Doozer]

I’m waiting until ChatGPT5 any predictions as to what capabilities it could have? I am guessing it will be less repetitive and finally upgrade the cutoff-date.

[funkervogt]

So I think I agree with Sam Altman, we could be at human level AGI in five years or something from now. I also won't be shocked if it's 15 years, I'll be shocked if it's 50 years. And what's cool is how uncontroversial these statements are in so many circles are now, right? It's cool and it's scary,

[wjfox]

[wjfox]

Guys, can you please stop creating all these new AGI threads. Keep stuff in this main thread. Thanks.

[weatheriscool]

Timeline of Open and Proprietary Large Language Models
April 5, 2023 by Brian Wang

Here is a timeline of open and proprietary large language models.


https://www.nextbigfuture.com/2023/04/t ... odels.html

[weatheriscool]

India Says It Won't Regulate AI
3
Mack DeGeurin
https://news.yahoo.com/india-says-wont- ... p_catchall

In a statement on Tuesday, India’s Ministry of Electronics and Information Technology acknowledged numerous ethical concerns around bias and transparency that could arise with AI’s rapid expansion but explicitly said the Indian government “is not considering bringing a law or regulating the growth of artificial intelligence in the country.” The ministry instead referred to AI as a “kinetic enabler of the digital economy,” which it believes will strengthen entrepreneurship and business and play an important strategic role for the country moving forward.

The ministry did say that officials were working to standardize responsible AI guidelines to steer AI’s development and promote healthy growth in the industry but notably shied away from expressing the same level of alarm raised by a growing number of policymakers in the US and Europe.

[weatheriscool]

Open Source Vicuna Claims 90% Quality VS GPT-4
April 7, 2023 by Brian Wang
https://www.nextbigfuture.com/2023/04/o ... gpt-4.html

Vicuna-13B, an open-source chatbot trained by fine-tuning LLaMA on user-shared conversations collected from ShareGPT. Preliminary evaluation using GPT-4 as a judge shows Vicuna-13B achieves more than 90%* quality of OpenAI ChatGPT and Google Bard while outperforming other models like LLaMA and Stanford Alpaca in more than 90%* of cases. The cost of training Vicuna-13B is around $300. The training and serving code are publicly available for non-commercial use.

The Vicuna code is at FastChat’s Github.
https://vicuna.lmsys.org/

There is an online demo here

Vicuna-13B is trained on user-shared conversations from ShareGPT and fine-tuned on LLaMA. It is backed by an enhanced dataset and an easy-to-use, scalable infrastructure.

[spryfusion]

[Powers]

Open Source Vicuna Claims 90% Quality VS GPT-4
April 7, 2023 by Brian Wang
https://www.nextbigfuture.com/2023/04/o ... gpt-4.html

Preliminary evaluation using GPT-4 as a judge shows...

I wonder how that worked out.

[weatheriscool]

I am finding that Claude is going more indepth in explaining science and stuff to me. I find it awesome. Chat-GPT is awesome too.

I really liked Claudes version of the mega snake story! It brought more passion!

[weatheriscool]

Six Big AI Projects in the Race Beyond 2 Trillion Parameters
April 8, 2023 by Brian Wang

https://www.nextbigfuture.com/2023/04/s ... eters.html

There are six major AI projects in the race beyond 2 trillion parameter models. There is OpenAI, Anthropic, Google/Deepmind, Meta, a UK government project and a stealth project.

It takes $1-2 billion per year of resources to be in the game. There is $1 billion needed for hardware and that hardware will need to be updated every 2-3 years. There is a need for hundreds of AI specialists and staff and some of the lead people will need $1-2 million and stock options.

There are over a hundred other projects and some of those could step up and impact the race.

GPT-5 should be finished by the of 2023 and released early in 2024. It should have 2-5 trillion parameters.

Anthropic plans to build a model called Claude-Next which should be 10X more capable than today’s most powerful AI. Anthropic has already raised $1 billion and will raise another $5 billion and will spend $1 Billion over the next 18 months. Anthropic estimates its frontier model will require on the order of 10^25 FLOPs, or floating point operations. This is several orders of magnitude larger than even the biggest models today. Anthropic relies on clusters with “tens of thousands of GPUs.” Google is one of the Anthropic funders.

[weatheriscool]

Vicuna is the Current Best Open Source AI Model for Local Computer Installation
April 9, 2023 by Brian Wang

Vicuna is a new, powerful model based on LLaMa, and trained with GPT-4. Vicuna boasts “90%* quality of OpenAI ChatGPT and Google Bard”. This is unseen quality and performance, all on your computer and offline.



The way I am reading this is 1. It is open source, 2. Can be downloaded onto your computer and 3. Offline.

This makes it a hell of a lot harder to stop by italy. lol. Or what ever.

In the future everyone could have a copy of a.i and the black market in some places could get rich in trade for a transfer on the black web or a simply passing by hand of a harddrive or even a c.d.

[weatheriscool]

Powerful new Meta AI tool can identify individual items within images
https://techxplore.com/news/2023-04-pow ... idual.html


by Peter Grad , Tech Xplore

Meta took a big leap forward this week with the unveiling of a model that can detect and isolate objects in an image even if it never saw them before. The technology is introduced and described in an article on the arXiv pre-print server.

The AI tool represents a major advance in one of technology's tougher challenges: allowing computers to detect and comprehend the elements of a previously unseen image and isolate them for user interaction.

It recalls a concept the former chair of the National Security Commission on Artificial Intelligence Robert O. Work once described: "What AI and machine learning allows you to do is find the needle in the haystack."

In this instance, Meta's Segment Anything Model (SAM) hunts for related pixels in an image and identifies the common components that make up all the pieces of the picture.

"SAM has learned a general notion of what objects are, and it can generate masks for any object in any image or any video, even including objects and image types that it had not encountered during training," Meta AI announced in a blog post Wednesday.

The recognition task is called segmentation. We do it daily without a moment's thought. We recognize items on our offices desks such as smartphones, cables, computer screen, a lamp, a melting candy bar, a cup of coffee.

[weatheriscool]

Alibaba to roll out its rival to ChatGPT across all its products
Published Mon, Apr 10 202310:01 PM EDTUpdated 9 Min Ago
https://www.cnbc.com/2023/04/11/alibaba ... ducts.html

Alibaba Cloud, the cloud computing unit of Chinese tech giant Alibaba, announced Tuesday it will be rolling out its own ChatGPT-style product Tongyi Qianwen.

Tongyi Qianwen, which possess Chinese and English language capabilities, will initially be deployed on DingTalk, Alibaba’s workplace communication software, and Tmall Genie, a provider of smart home appliances, the company said in a release.

At the 2023 Alibaba Cloud Summit, the company said it will be rolling out the artificial intelligence-powered chatbot into all Alibaba products from enterprise communication to e-commerce in “the near future.” It did not reveal a timeline.

“We are at a technological watershed moment driven by generative AI and cloud computing, and businesses across all sectors have started to embrace intelligence transformation to stay ahead of the game,” said Daniel Zhang, chairman and CEO of Alibaba Group and CEO of Alibaba Cloud Intelligence, in a statement.

[wjfox]

[wjfox]

[weatheriscool]

Amazon Is Joining the Generative AI Race
The ecommerce giant doesn’t have a ChatGPT rival, but it wants to sell you the tools you need to build one.
https://www.wired.com/story/amazon-is-j ... e-ai-race/

Amazon may still be king of the cloud, but in recent months it’s had to watch its two closest rivals, Microsoft and Google, steal the spotlight with brilliant but error-prone chatbots that use cutting-edge “generative” artificial intelligence models.

Today, Amazon announced it’s joining the generative AI race. Not by launching its own chatbot, but by making two new AI language models available through its cloud platform, Amazon Web Services, which customers will be able to use to build their own bots.

The past few months have seen the tech industry foaming at the mouth over the potential of generative AI—algorithms that learn to produce text, code, imagery, and more. The boom has been inspired by the remarkable success of OpenAI’s text-generating bot, ChatGPT, as well as the success of AI-image generators.

[wjfox]

[weatheriscool]

[Yuli Ban]