Google’s AI tool Gemini takes a bashing on “woke” accusations

AI is a major talking point in the agency at the moment and the cultural bias that large language programmes demonstrate is proving a real challenge to AI-generated content and imagery.

Gemini (Google’s AI tool), who has built into the algorithm a way to generate more diverse responses, has experienced backlash and been accused of being too “woke” – Viral posts have been circulated of Gemini’s artificially generated image from the prompt  “image of the US Founding Fathers which inaccurately included a black man”.

Google responded rapidly by pausing the tool and highlighting in a blog post that Gemini “missed the mark”. However, the hyper-vigilant politically correct images still were being generated and displayed to users.

Read more here.

Meta trails cross-posting test between Facebook and Threads

This week, Meta introduced a cross-posting trial between Facebook and Threads, enabling selected users to share Facebook updates on the Twitter-like app. Limited to iOS, the feature covers text and link posts, excluding videos, and is not tested in the EU. 

While streamlining sharing, questions arise about user preferences for identical content across platforms. Despite Meta’s optimism, maintaining Threads’ distinct identity amid integration challenges remains crucial for its growth.

Read more here.

X job listings surge: A new way to advertise vacancies?

In recent weeks, X has made strides with its latest feature known as ‘X Hiring’, boasting over a million job opportunities exclusively for X Premium subscribers. Initially rolled out on business profiles last August, job postings were accessible across all X Premium packages, including personal profiles for US subscribers.

The surge in job listings, from 750,000 to over a million, indicated growing utilisation. However, while the increasing number of listings suggested a rise in popularity, it is crucial to evaluate their effectiveness. Experts have claimed that Insights into how many roles were filled via X Hiring would provide a clearer picture, with many questioning the validity of certain job listings and their functionality.

Read more here.

Supreme Court debates social media regulation

This week, the US Supreme Court deliberated pivotal cases on social media regulation, potentially reshaping the digital landscape. Texas and Florida seek increased control over platforms like Facebook and TikTok, prompting debate on the authority of platforms versus regulation from different states.

The dispute centres on whether platforms can dictate content independently or if states can regulate removal policies. A ruling favouring the states could impact the 2024 elections’ information sharing, with significant implications for online free speech rights.

Read more here

Meta combat ‘deceptive AI’ ahead of EU elections

This week, Meta pledged to tackle deceptive AI content before the EU elections in June, wary of potential manipulation of voters through generative AI. Despite doubts about its effectiveness, Meta set up an internal Operations Centre for the EU, so that they can swiftly address emerging threats.

Meta stressed the importance of collaboration with fact-checking bodies and has since enforced strict ad guidelines to combat misinformation. Meta’s proactive approach highlighted the necessity for industry-wide collaboration in protecting misinformation in politics.

Read more here.