AI WEEKLY NEWS (14) – WEEK 12-August to 18-August-2023:

This article is not generated by AI.

亚马逊 announced that it will use AI to enhance customer reviews

[About Amazon]

The Allen Institute has published a 3 trillion tokens dataset that’s free to use and open to inspection

Dolma is an open dataset of 3 trillion tokens from a diverse mix of web content, academic publications, code, books, and encyclopedic materials.?[github]

Anthropic is raising more cash – this time $100M from a South Korean Telcom

[SK Telcom]

NYT is considering legal action against OpenAI

Both parties are in the middle of licensing discussions about including 纽约时报 articles in OpenAI tools and those discussions seem to have got stuck.

So, how about a Spanish Inquisition approach – showing the instruments.?Crazy...

Policy makers will have to soon assure a base on which AI companies can securely work.?If it's put on the internet to read, then fair use with AI must be allowed or excluded explicitly. This of course, excluding explicitly, will most likely include indexing by search engines as most of them will / are using AI to process such content.??

Publishers or content creators can reserve rights forward looking or change the terms of use for material they published in the past. But claiming damages for what they published in the past without such terms is ludicrous. And if content created by AI can’t be protected by copyright laws, we have to also tightly limit cases where such use would infringe the rights of others, e.g. direct copies, reproductions, etc.

We are going to a new way to use content on the internet. Soon your personal AI assistant will filter out all the click-bait or nonsense / ad only articles. For the benefit of the content creators and their audience, the content / the Internet needs to be available for processing in many ways, it must be possible to find and process it without the threat of a law suite. Or otherwise, reading an article on the internet might soon constitute a copyright infringement.

[NPR]

Gartner places Generative AI in their Hype cycle on “Peak of inflated expectations”

[Gartner]

Australian News Corp produces 3’000 news stories a week with AI

[The Guardian]

Associated Press puts up standards around Generative AI to stop staff from using AI to write publishable content

"...It is the responsibility of AP journalists to be accountable for the accuracy and fairness of the information we share..."

[AP Blog]

#ai #artificialintelligence #machinelearning #OpenAI?

#ainews #generativeai #tech #deeplearning #llm #llms #genai?

Please support us by liking and re-posting our Weekly AI News articles!