Musk Says He’ll Drop $97b Bid for OpenAI if the Board Stop For-Profit Transition Talks

⚡️ Headlines

🤖 AI

Adobe launches public beta for AI-generated video feature – Adobe debuts its AI video generation tool, allowing users to create video content from text prompts. [The Verge]

New hack corrupts Gemini AI's long-term memory via prompt injection – Researchers reveal a security flaw that allows attackers to manipulate AI's memory by injecting malicious prompts. [Ars Technica]

OpenAI to release O3 Deep Research Agent for free and ChatGPT Plus users – OpenAI CEO Sam Altman outlines plans to make the powerful O3 AI assistant available to a broader user base. [VentureBeat]

OpenAI faces criticism for handling controversial topics in ChatGPT model – OpenAI's moderation decisions on sensitive topics spark debate on bias and transparency in AI models. [The Verge]

OpenAI to offer free users unlimited access to GPT-5 – OpenAI plans to extend GPT-5 access to all ChatGPT users, removing limitations on free-tier usage. [Engadget]

AI chip startup Groq secures $1.5 billion from Saudi Arabia – Groq lands major funding from Saudi Arabia to advance its AI hardware technology. [Reuters]

🤳 Social Media

YouTube outlines major plans for 2025, focusing on AI tools and creator monetization – YouTube shares its vision for AI-enhanced discovery, expanded monetization options, and new creator support. [YouTube Blog]

Reddit reports strong growth in 2024, with focus on advertising and community engagement – Reddit highlights impressive annual results, driven by ad revenue and active user growth. [Stock Titan]

⚖ Legal

Russia and U.S. discuss possible prisoner exchange involving Marc Fogel – Talks are underway for a potential exchange to secure the release of Marc Fogel from Russian detention. [The New York Times]

Musk's X agrees to pay $10 million to settle Trump lawsuit – Social media platform X settles a lawsuit related to former President Trump's account restrictions. [Wall Street Journal]

EU scraps AI liability directive after criticism from Vance – The European Commission withdraws proposed AI regulation following backlash from industry leaders. [Euractiv]

🎱 Random

Trump plans to nominate Brian Quintenz as head of U.S. derivatives regulator – The former CFTC commissioner is expected to lead the agency under Trump's administration. [Bloomberg]

🔌 Plug-Into-This

Elon Musk States He Will Withdraw $97 Billion Bid for OpenAI if they Stay Non-Profit

Elon Musk has announced he will withdraw his $97.4 billion offer to acquire OpenAI if the organization decides to remain a nonprofit and halts its conversion to a for-profit entity.

• Musk's lawyers submitted a filing to a California court stating that if OpenAI's Board commits to maintaining its mission and stops its for-profit conversion, Musk will rescind his bid.

• OpenAI CEO Sam Altman has already rejected the unsolicited bid, affirming that the company is not for sale.

• OpenAI, originally founded by Musk and others in 2015, is controlled by a nonprofit board with a stated mission of safely developing superior AI for public benefit.

🤑 No surprise here, we knew it was effectively the reason for the bid. Nobody seems to think Elon is serious about buying the company, but his aim to put pressure on the board and complicate their impending transition to a for-profit entity has been clear from the start. It still seems unlikely to actually work, but he has probably succeeded in sowing some discontent within the OpenAI ranks, as the valuation number (which he’s putting at $97b) basically determines the compensation for the board when the non-profit does disband.

DeepSeek Gives China's Chipmakers a Competitive Edge in AI

DeepSeek's AI models are helping Chinese chipmakers like Huawei improve their competitiveness in AI inference, closing part of the gap with U.S. firms such as Nvidia. DeepSeek’s focus on computational efficiency and its open-source nature are boosting adoption in China, especially amid U.S. export restrictions on high-end AI chips.

• Huawei and other firms are integrating DeepSeek into AI inference tasks, avoiding dependence on Nvidia’s GPUs.

• U.S. restrictions block advanced Nvidia chips, but less powerful versions remain available for Chinese inference use.

• Nvidia's CUDA platform still dominates, making software compatibility a hurdle for Chinese firms.

• Huawei's CANN framework aims to rival CUDA but faces developer resistance.

• DeepSeek aligns well with industry-specific applications, strengthening local innovation.

🐳 DeepSeek’s rise aligns with China's strategic shift toward homegrown AI solutions and reduced reliance on U.S. tech, which appears to be working well so far.

OpenAI Cancels O3 AI Model in Favor of Unified Next-Gen Release

OpenAI has decided to cancel the release of its anticipated O3 model, shifting focus toward a more comprehensive next-generation AI release. The company aims to deliver a unified model that merges advancements across various AI capabilities, offering enhanced performance and versatility.

• The unified model will combine language, vision, and reasoning improvements.

• OpenAI's shift aligns with its broader goal of consolidating AI capabilities into fewer, more powerful models.

• The cancellation avoids fragmenting its development pipeline.

• Users can expect a significantly upgraded experience in the next major release.

• The timing reflects OpenAI’s strategy to stay competitive amid accelerating AI advancements.

🧠 There has already been a lot of anticipation for GPT-5 and o3 separately, and since DeepSeek’s arrival, OpenAI appears to be increasingly under pressure from without and within to produce massive improvements with every release. That’s hard to do no matter who you are, so we can probably read this as a move to avoid small releases that flop and fuel more public ridicule.

 🆕 Updates

📽️ Daily Demo

🗣️ Discourse