😎 Orion, Meta announces AR headset

ALSO: Meta’s Connect details and OpenAI loses more key staff

Estimated Read Time: 5 minutes

Meta just had their Connect event, where they detailed their plans for tech and AI. Surprisingly they announced a new AR headset (think a competitor to Apple’s Vision Pro) and new features coming to Facebook and Instagram using AI. Plus OpenAI has more drama as the company prepares to become for-profit.

• 🤖 Meta’s Event, Meta AI and Llama get upgrades.

• 🔥 Several key staff leave OpenAI.

• 🎥 Summarize Youtube videos.

• 💰 Salesforce acquires AI company Zoomin.

🤖 Meta’s Event, Meta AI and Llama get upgrades

What happened: Meta Connect just ended, resulting in several big tech and AI announcements made by the lizard man himself. Including AR glasses which actually look high quality and easy to use.

The details: 

• Zuckerberg announced Orion, Meta’s prototype for its first consumer full holographic AR glasses. Main selling points are the lightweight design along with hand and eye tracking. No release date was announced.

• The Quest 3S VR headset is launching October 15th at $299 for the base model. It’s entirely wireless and supports existing Quest apps and games. The older Quest models will be phased out until they run out of existing inventory.

• Meta AI has a voice now. You can converse with Meta AI out loud through Messenger, Facebook, WhatsApp and Instagram. And it’ll talk back in one of multiple voice options modeled after celebrities.

• Meta AI has visual search and enhanced image editing. Now you can interact with images by prompting the AI to respond to questions or edit photos based on feedback. You can post edited images straight to Instagram.

• Meta AI has 500 million users, according to Zuckerberg.

• Llama’s been upgraded to 3.2, introducing multimodal capabilities (it can recognize pictures, analyze maps, look at graphs). It comes in 11B and 90B parameter models. But it’s not available in EU regions because of regulations.

• The Ray-Ban Smart Glasses now have AI that allows users to ask questions about their surroundings and receive immediate feedback. Other new features include a reminder system, live language translation for supported languages, and music streaming (from certain services like Amazon and iHeartRadio).

• Facebook now suggests AI generated content it thinks the user will like.

• Facebook is also testing AI powered translations and dubs for videos. The first supported languages will be English and Spanish.

Why it matters: Meta’s plans for AI and tech are massive. How AI is used in social media is going to affect the daily life of most people. I think Facebook’s approach of integrating AI while making sure users know it isn’t real is the best. I’m more wary of the headsets though. But with enough quality of life, the AR and VR headsets might push their way to mainstream popularity.

🧠 Microsoft has introduced “Correction,” a service which detects and fixs factual inaccuracies in AI-generated text by cross-referencing it with reliable sources. Apparently experts are skeptical, suggesting that it may not resolve the underlying issue of AI hallucinations and could potentially create new problems. Additionally, there are concerns about the feature possibly giving users a false sense of security and the business motives behind its implementation. This is just another example of AI companies trying to rid AI of its biggest problem: hallucinations.

💰 Letta, founded by Berkeley PhD students, emerged from stealth with $10 million in seed funding to commercialize MemGPT—an open-source project that helps AI models remember users and conversations. They position themselves as an alternative to OpenAI, Letta plans to offer Letta Cloud, a hosted service enabling developers to deploy stateful AI agents compatible with multiple AI models. They’re worth keeping an eye on.

🔥 Several key staff leave OpenAI

What happened: Multiple key staff from OpenAI just left as the company prepares to become a traditional for-profit company.

The details:

• OpenAI CTO (Chief Technology Officer) Mira Murati announced her departure from the company after six years.

• Murati's decision comes amid Sam Altman's increasing control over OpenAI and its image.

• Murati led OpenAI's work on ChatGPT, DALL-E, and Codex.

• Sam Altman is reportedly about to receive equity (money, basically) in the company for the first time. This happens as OpenAI is setup to become a more traditional for-profit company next year.

• Then their chief research officer, Bob McGrew, and research VP, Barret Zoph, resigned after CTO Mira Murati announced her departure.

• Mark Chen is promoted to SVP of research, Matt Knight becomes chief information security officer, Kevin Weil and Srinivas Narayanan continue to lead the applied team, and Josh Achiam becomes head of mission alignment.

Takeaway: Ironically ChatGPT will refuse to answer any questions or write content about this topic. But clearly there’s come conflict brewing in the company, and has been for a while. The company has been chasing progress and profit at the sake of safety for the last year. Thankfully there’s plenty of advanced competitors in the market, some of which consist of former OpenAI employees.

🎥 Summarize Youtube videos

What happened: Google has announced significant updates to its AI-powered note-taking and research assistant, NotebookLM. The tool now supports summarizing YouTube videos and audio files, and introduces the ability to create and share AI-generated audio discussions. These enhancements aim to broaden NotebookLM’s use cases and appeal to a wider range of users beyond educators and learners.

The details: 

• NotebookLM can now generate key summaries from Youtube videos and audio recordings.

• Here’s a neat feature: You can create Audio Overviews of your documents and share them via public URLs. 

• Google insists the info uploaded to NotebookLLM is private and not used to train their AI models.

• Mobile apps for NotebookLLM are expected to launch next year.

Why this matters: Summarizing and note taking are the most basic AI functions. This is a taste of how much people will use AI in their daily lives. Also incredibly useful to know if you learn from Youtube videos often.

💰 Salesforce acquires AI company Zoomin

What happened: Salesforce has announced plans to acquire Zoomin, an enterprise knowledge platform that unifies company documentation—such as product user guides and tutorials—into a single portal. Basically an AI and data focused acquisition.

The details:

• The sales price wasn’t disclosed. Salesforce expects the deal to be finalized in the fourth quarter of its fiscal year 2025 (by January 31).

• Zoomin utilizes AI and data to build self-service documentation for big companies. This is what companies need most at the moment.

• Current clients include McAfee and Dell, as well as fast-food franchises such as Burger King, Tim Hortons, and Popeyes. Even fast food needs their data documentation.

• Zoomin has previously raised $73 million in funding.

Takeaway: Salesforce is a major players in the corporate world. If you work in that environment, checkout what Salesforce is doing with AI. That’ll shape the direction companies go in the future. Right now it seems to be about organizing large swathes of raw data using AI.

🎨 Cool artstyles!

Source: Midjourney

Give us your feedback!

Got feedback? Rate today’s newsletter by clicking below! 

Got more personal feedback? Fill in the box when you rate us and give us your criticism and feedback.

Thank you for reading! 

❤️Share the (ARTIFICIAL) Love!

Got a friend who wants to appear as smart as you? An AI fanatic or someone shaking in their boots of getting replaced? Share us by clicking the link below and you’ll get our love ;)