In today’s newsletter:

• News - 🚨 Major Shakeup at OpenAI: Key Leaders Depart

• Productivity - 🎧 Turn Any Article or Paper into a Podcast with Rome’s AI!

• Culture - 🏈 Tom Brady Joins the Viral “Goodbye Meta AI” Trend—But Does It Work?

🚨 Major Shakeup at OpenAI: Key Leaders Depart

OpenAI has seen a major shakeup in its leadership, with the departure of Chief Research Officer Bob McGrew, VP of Research Barret Zoph, and CTO Mira Murati, all announced within hours of each other. While CEO Sam Altman emphasized that these exits were independent decisions and aimed to ensure a smooth transition, the timing is notable as OpenAI is undergoing rapid growth. Altman announced that Mark Chen will step up as SVP of Research alongside Jakub Pachocki as Chief Scientist, with several other internal promotions to stabilize the company as it continues its mission to push AI innovation.

Despite assurances from McGrew and Zoph that their departures were amicable and part of their personal career decisions, these exits follow a series of high-profile resignations from OpenAI in recent months, raising questions about internal dynamics. The changes come as OpenAI is reportedly transitioning from a nonprofit-governed organization to a for-profit entity, potentially leading to tension over the company’s direction. As OpenAI gears up for an all-hands meeting, there’s speculation about the long-term impact of these shifts, especially given OpenAI's crucial role in the fast-evolving AI industry.

🗞️ In Other News…

• Meta's annual developers conference shifts focus from VR to AI, introducing Meta AI with celebrity voices and smart glasses with real-time translation, showcasing a strategic pivot towards generative AI despite previous VR underperformance.

• California's SB 1047, requiring AI model safety tests, splits tech and business leaders; some see it as premature and potentially stifling, while others view it as a necessary step toward safer AI usage. Governor Newsom has until Sept. 30 to decide its fate.

• Tech giants' AI ambitions drive up their carbon emissions, contradicting earlier pledges to cut them, as massive energy needs for AI data centers spur collaborations with power producers to develop clean energy solutions like geothermal and nuclear power.

• Puma partners with Google Cloud to use generative AI for faster, personalized online branding, seeing a 10x speedup in image creation and improved engagement in markets like India.

• AMD's CEO Lisa Su forecasts a transition towards more specialized, energy-efficient chips as AI model designs standardize, reducing the need for the programmability dominant in today's GPUs.

🎧 Turn Any Article or Paper into a Podcast with Rome’s AI!

Rome’s Knowledge AI podcast tool works by allowing users to input a URL related to any topic they’re interested in. Once the link is provided, the AI scans the content, performs research on subtopics, and gathers key information. It then crafts this data into a cohesive and engaging podcast episode, which can be listened to on the go.

This system simplifies the process of turning written content into audio, making it easy for users to consume knowledge in an audio format while driving, exercising, or relaxing. Whether you want to dive into complex topics like neural networks or explore cultural histories, Rome turns web content into an easy-to-digest, personalized podcast just by inputting an article or paper.

🏈 Tom Brady Joins the Viral “Goodbye Meta AI” Trend—But Does It Work?

The “Goodbye Meta AI” message, which claims to prevent Meta from using Facebook, Instagram, and WhatsApp accounts for AI training, has gone viral, shared by celebrities like James McAvoy, Ashley Tisdale, and Tom Brady. Many users are posting it in an attempt to protect their privacy, believing that it asserts their right to opt out of Meta's controversial AI training practices. However, despite its widespread circulation, this message has no legal effect and doesn't stop Meta from using publicly shared content for AI purposes.

Meta has clarified that posting this message doesn’t count as a valid objection, similar to previous privacy “disclaimers” that periodically surface online. These posts stem from a longstanding misconception that a simple statement can protect users from legal or privacy concerns. In reality, users agree to Meta’s terms and conditions when they create an account, and the only effective way to opt out of having content used for AI training is by making accounts private, not by posting viral copypasta.

Job Opportunities

• Anyscale: Research Engineer

• Shield AI: Technical Program Manager

• Shield AI: Product Manager

• Mistral AI: Account Executive, Digital Natives (Germany)

• Databricks: Emerging Enterprise Account Executive

VC Fundraising Rounds

• Mediwhale, an AI-powered health diagnostics company, announced it has secured $12 million in Series A2 funding. (9/26/24)

• Supabase Inc., an open-source application development platform and Postgres database service provider, announced it has secured $80 million in new funding, bringing its total raised to $196 million. (9/25/24)

• Vsim, a Manchester-based startup developing a new physics simulation framework, has raised $21.5 million in seed funding. (9/25/24)

• GEOX, an Israeli startup that created an AI-driven platform to assess real estate risk from climate-related disasters, has secured $19 million in Series A funding. (9/25/24)

• Tamnoon Inc., a managed cloud security remediation startup, secured $12 million in new funding to advance its product development, grow its partner network, and boost innovation in cloud security remediation services. (9/25/24)

Earn Free Rewards 🎁 

Get rewarded for helping others level up their AI game 👇️

25 referrals - 30 minute chat with Ben Chon, founder of rareliquid

1 referral - Daily VC AI Fundraising Tracker

You currently have 0 referrals.

Copy & paste this link: https://www.rarecandyai.com/subscribe?ref=PLACEHOLDER