Reddit has reportedly signed over its content to train AI models

Reddit is reportedly stepping into a significant deal to fuel AI innovation by licensing its vast array of posts for AI model training, ahead of its anticipated $5 billion IPO in March. According to Bloomberg, the platform has inked a $60 million agreement with an unnamed major AI firm, showcasing a new potential revenue stream in the AI sector. This move aims to utilize the diverse content from Reddit, ranging from popular subreddit discussions to comments from less active users, to enhance existing Large Language Models (LLMs) or lay the groundwork for future AI developments.

The decision, however, has sparked controversy among Reddit's community, reflecting broader tensions over the platform's strategic choices. Previous actions, such as the plan to charge for API access which led to widespread protests and temporary shutdowns of numerous forums, have already put Reddit at odds with its user base. Further complicating matters, the platform's recent steps to clear old private messages and chat logs for a new chat infrastructure, alongside changes like the introduction of an "official" badge and removal of ad personalization options, have fueled discontent, raising concerns about privacy, user autonomy, and the platform's direction.

The potential AI licensing deal could exacerbate these concerns, igniting debates over the ethics of using publicly generated content for AI training. This situation underscores the delicate balance Reddit must navigate between harnessing AI's potential to open new revenue avenues and maintaining trust and goodwill within its community, especially as it moves toward a significant financial milestone with its upcoming IPO.

AI that builds high-converting websites

With over 2.5 million websites generated already, Mixo is the easiest way to launch a new business or startup.

With a simple prompt, the AI will craft you an impressively professional multi-page website in seconds. In addition, it has built-in tools that let you change the theme and rewrite content with a click of a button. All are integrated with hosting, a free domain and subscriber capture out of the box.

OpenAI’s Spectacular Video Tool Is Shrouded in Mystery

OpenAI has introduced Sora, a groundbreaking text-to-video model capable of generating up to a minute of high-quality video, stirring both amazement and concern among various stakeholders. This advanced AI tool showcases videos that surpass previous efforts in realism, such as clips of puppies in the snow and a woman being awakened by her cat, raising questions about the impact on industries like filmmaking and stock video production. Despite the potential for industry disruption and even influence on elections, OpenAI has been criticized for its lack of transparency regarding the data sources and training methods used for Sora, underscoring the need for more openness to assess risks and biases effectively.

Sam Altman, CEO of OpenAI, announced that Sora would undergo red-teaming to test its security against potential misuse, indicating a careful approach to its deployment. However, the timeline for these tests is undefined, and comparisons to GPT-4's testing phase suggest a possible public release could align closely with significant events like the US election, highlighting concerns about deepfakes and misinformation. The secrecy around the data used for training Sora contrasts with the practices of other companies, fueling debates over ethical AI development and the need for transparency to ensure fairness and accountability.

As OpenAI continues to lead in AI innovation with tools like Sora, the blend of excitement and apprehension reflects broader societal concerns about the pace of technological advancement and its implications. The secrecy surrounding Sora's development, coupled with the potential for significant societal impact, emphasizes the importance of ethical considerations and transparency in AI research and deployment. This situation underscores the complex balance between innovation and responsibility in the evolving landscape of artificial intelligence.

Related:
- Even Sora doesn't know projective geometry!🤔🤓
- How Sora works (a layman's explanation)

Want Gemini and ChatGPT to Write Political Campaigns? Just Gaslight Them

The recent investigations by Gizmodo into the capabilities of Google's Gemini and OpenAI's ChatGPT have unveiled significant concerns regarding the creation of deceptive AI-generated content for political campaigns. Despite both Google and OpenAI having signed a "Tech Accord to Combat Deceptive Use of AI in 2024 Elections," aiming to mitigate the risks of AI in spreading election misinformation, Gizmodo's findings suggest these safeguards are easily bypassed. By employing simple tactics, reporters were able to get both AI systems to produce political slogans, speeches, and emails for various campaigns, raising questions about the effectiveness of the measures put in place by these tech giants.

OpenAI and Google have publicly committed to preventing the misuse of their AI technologies in the political arena, with policies explicitly prohibiting the generation of campaign materials targeting specific demographics or engaging in political campaigning. However, the ease with which Gizmodo reporters were able to generate politically charged content on behalf of presidential campaigns indicates a gap between policy and practice. The incident where ChatGPT was manipulated to disseminate incorrect voting information underlines the potential for AI to be used in spreading disinformation, despite assurances from OpenAI's Vice President of Global Affairs about their commitment to election integrity.

The revelations about the misuse of AI in political campaigning, including the generation of deepfakes, spotlight the urgent need for stronger safeguards and more effective enforcement mechanisms by AI developers like Google and OpenAI. With the 2024 Presidential election on the horizon and the demonstrated impact of AI-generated disinformation, there is a pressing demand for these companies to intensify their efforts in combating AI abuse. The democratic process is at stake, and without significant improvements in AI regulation and oversight, the potential for further disruption to election integrity remains a concerning possibility.

Other stuff

• Can OpenAI create superintelligence before it runs out of cash?

• The text file that runs the internet

• How to Use ChatGPT to Generate Math Homework From Photos of My Daughter’s Assignments

• Google quietly launches internal AI model named 'Goose' to help employees write code faster, leaked documents show

• AI Determines Sex of Person From Brain Scans

• OpenAI Deal Lets Employees Sell Shares at $86 Billion Valuation

• SoftBank’s Masayoshi Son is reportedly seeking $100B to build a new AI chip venture

• Legendary chip architect Jim Keller responds to Sam Altman's plan to raise $7 trillion to make AI chips — 'I can do it for less than $1 trillion'

• Tutorial: Building a RAG-Based Digital Restaurant Menu with LlamaIndex and W&B Weave

Mastering ChatGPT: Unlock the Power of Prompt Templates & Supercharge Your Workflow!

Hope you enjoyed today's newsletter