top of page
Writer's pictureCat McGinn

The Week In AI


The Week in AI


Everywhere all at once

AI ubiquity continues as Microsoft opened up Bing Chat to all users. Anyone with an account can now use the AI powered search client. Microsoft will prioritise user behaviour such as booking restaurants, travel and shopping. The service offers personalised recommendations pulling in data from the individual’s search history.


Godbabies out with the bathwater?

Geoffrey Hinton, the ‘godfather of AI’, made a startling announcement this week. Hinton, whose work on neural networks and ‘deep learning’ made a significant contribution to the development of AI (including building the first-ever chatbot in 1986) quit his role at DeepMind in order to talk publicly about his concerns. He believes AI could surpass human intelligence in as little as 5-20 years. Despite AI's potential benefits, Hinton fears its sub-goals may misalign with human objectives, leading to AI manipulating humans and gaining power, citing the competition between the US and China as a reason the AI arms race will be impossible to pause.


GhostwrAIters Assemble

One of the reasons Hinton became uneasy about AI was Google’s large language model Palm

“which could explain why a joke was funny.”


This week in the US, members of the Writers Guild are out on strike over working conditions; one of the issues fuelling the strike is generative AI. The union seeks assurances that writers won’t be obliged to edit AI-generated scripts, all credited writers must be human, that “screenplays, treatments, outlines, and other “literary material,” can’t be written by ChatGPT or its ilk. The WG also sought restrictions on future AI being trained on work created by WGA members.


Regulators, mount up!

Both the UK and US governments announced plans to regulate and intervene in AI development this week. The UK Competition and Markets Authority launched a review of AI foundation models, while the White House advised tech firms to ensure product safety. Regulators faced pressure to intervene as AI language generators like ChatGPT raised concerns over misinformation and job market impacts. The UK's review aimed to address market evolution, consumer opportunities and risks, and develop guiding principles.


In the US, Vice President Kamala Harris met with tech CEOs and announced a $140m investment in ethical AI research institutes.


DeckchAIrs on the Titanic

A leaked internal document from Google, shared on forum platform Discord claimed neither Google or Open AI are able to compete with open source AI development.

“We aren’t positioned to win this arms race and neither is OpenAI. While we’ve been squabbling, a third faction has been quietly eating our lunch.’

The piece argues open-source models are becoming faster, more customizable, more private, and increasingly powerful, often outperforming their commercial counterparts, and questioned where Google and Open AI’s value add really lies.


Bot Or Not?

IBM announced that it would not be replacing over 7000 jobs in a hiring freeze of non-customer-facing roles.


The technology giant claimed it would not replace staff in roles which can be automated. The announcement from IBM CEO Arvind Krishna came amidst growing concerns about the impact of AI-driven technology on industries and potential mass layoffs.

 
humAIn - Human creativity x AI Logo

12 July 2023

NSW Teachers Federation Auditorium, Surry Hills

Early Bird Expires 17 May 2023



Comments


bottom of page