meta, amazon and anthropic Updates

Meta Platforms Inc. is significantly advancing its AI wearable strategy with the acquisition of Limitless, an AI startup formerly known as Rewind. Announced on Friday, December 5, 2025, the deal sees Limitless, known for its AI-powered pendant that records conversations and generates summaries, joining Meta's Reality Labs. Limitless CEO Dan Siroker stated the move aligns with Meta's vision for "personal superintelligence" through AI-enabled wearables, building on existing products like Ray-Ban Meta smart glasses. The acquisition, which follows Limitless having raised about $35 million, means Limitless will cease hardware sales, support existing customers for one year, and phase out its Rewind desktop software, with plans to integrate its technology with Meta's Llama AI model. Meanwhile, Amazon is intensifying its presence in the artificial intelligence sector, not only through a substantial $8 billion investment in Anthropic but also by developing its own advanced AI models. The company leverages its AWS cloud service, a critical advantage in the competitive AI sector. Amazon recently unveiled a suite of new Large Language Models, including Nova Light, Nova Pro, Nova Sonic for real-time voice applications, Nova Omni for experimental reasoning, and Nova Forge, a customizable LLM tailored for specific business requirements. In other significant developments, Google faces a legal setback as Judge Amit Mehta in Washington ruled that the company must limit its default contracts to one year. This decision, supporting the US Justice Department's antitrust efforts, prevents Google from paying companies like Apple and Samsung to exclusively set its search engine as the default on browsers and phones for periods longer than a year. Google plans to appeal the initial antitrust ruling. Separately, Goodwill of the Finger Lakes, in partnership with Google.org, is offering a free, six-hour Google AI Essentials online course until the end of 2026, aiming to equip individuals with generative AI skills for the workplace. Beyond the tech giants, a small startup named Poetiq made headlines by achieving 54% accuracy on the ARC-AGI-2 AI benchmark, outperforming Gemini 3 Pro at half the cost. Poetiq's "meta-system" approach cleverly utilizes existing large language models to improve accuracy and lower costs, and the company has open-sourced its code. The broader societal impact of AI continues to be debated, as seen in a Knight High School debate where students discussed AI's role in education, weighing benefits like 24-hour tutors against concerns over misinformation, privacy, and job displacement. The discussion around AI's impact on employment also saw conflicting views; while figures like Elon Musk and Bill Gates warn of widespread job elimination, LinkedIn's recent hiring data indicates an increasing demand for jobs, suggesting a more positive outlook, especially for Gen Z. However, investor Orlando Bravo cautioned at a Bloomberg event on December 6, 2025, that "fear of missing out" (FOMO) is driving hasty and potentially regrettable investments in AI tech firms. Even legal systems are grappling with AI's implications, as Ontario Councilor Corinna Traill, arrested for uttering threats, claimed an alleged threatening voicemail was "created using artificial intelligence."

Key Takeaways

• Meta acquired AI wearable startup Limitless on December 5, 2025, to boost its AI glasses and "personal superintelligence" vision.
• Limitless will stop selling its AI pendant and Rewind software, providing one year of support to existing customers.
• Meta plans to integrate Limitless's technology with its Llama AI model.
• Amazon invested $8 billion in Anthropic and introduced new Large Language Models (Nova Light, Pro, Sonic, Omni, Forge), leveraging its AWS cloud service.
• Google must limit its default search engine contracts with companies like Apple and Samsung to one year, following an antitrust ruling.
• Goodwill of the Finger Lakes, in partnership with Google.org, offers a free Google AI Essentials online course until the end of 2026.
• Startup Poetiq achieved 54% accuracy on the ARC-AGI-2 AI benchmark, outperforming Gemini 3 Pro at half the cost by using a "meta-system" with existing LLMs.
• LinkedIn data indicates an increasing demand for jobs, countering warnings from figures like Elon Musk and Bill Gates about AI-driven job elimination.
• Private equity investor Orlando Bravo warned on December 6, 2025, that AI "fear of missing out" (FOMO) is leading to hasty and potentially regrettable investments.
• Ontario Councilor Corinna Traill, charged with uttering threats, claimed an alleged threatening voicemail was "created using artificial intelligence."

Meta buys AI wearable company Limitless

Meta has bought Limitless, a company known for its AI-powered pendant that records conversations and creates summaries. Limitless will stop selling hardware and support existing customers for one year, allowing them to export or delete their data. This acquisition helps Meta expand its AI wearable strategy, building on products like Ray-Ban Meta smart glasses. Meta aims to integrate Limitless's expertise in always-on audio capture and processing into its own devices. The deal highlights Meta's focus on context-aware gadgets that understand the world around users.

Meta buys Limitless for AI wearable push

Meta acquired Limitless, a company known for its AI-powered pendant that records conversations and creates summaries. Limitless CEO Dan Siroker announced the deal on Friday, December 5, 2025, stating they share Meta's vision for "personal superintelligence" through AI-enabled wearables. This acquisition helps Meta speed up its work on AI glasses and other devices, a key focus for CEO Mark Zuckerberg. Meta is shifting resources towards AI, even considering budget cuts for its metaverse group. Ray-Ban Meta smart glasses, a partnership with EssilorLuxottica, already show strong sales growth.

Meta buys AI firm Limitless stops hardware sales

Meta announced it bought AI startup Limitless, formerly known as Rewind. Limitless will stop selling its AI wearable pendant and provide one year of support for current customers. Existing subscribers will move to an Unlimited plan, and the Rewind desktop software will also be phased out. Limitless co-founder Dan Siroker stated they support Meta's goal of bringing "personal superintelligence" to everyone through AI-enabled wearables. This move helps Meta strengthen its focus on AR/AI glasses like Ray-Ban Meta. Customers can export or delete their data during this transition.

Meta buys Limitless for AI wearable future

Meta has acquired Limitless, an AI wearable startup known for its pendant that records conversations and creates AI summaries. Limitless CEO Dan Siroker announced the deal on Friday, December 5, 2025, without sharing financial details. This acquisition fits Meta's vision of "personal superintelligence" through AI-enabled wearables and expands its hardware offerings beyond Ray-Ban smart glasses. Limitless's technology, which addresses challenges like battery life and usefulness, gives Meta a head start in the growing ambient AI market. Meta plans to integrate this technology into its ecosystem, potentially raising privacy questions about data collection.

Meta buys Limitless to boost AI smart glasses

Meta Platforms Inc. acquired Limitless, an AI startup known for its pendant that records conversations and provides insights. The deal, announced Friday, December 5, 2025, aims to advance Meta's AI hardware, especially smart glasses. Limitless's team will join Meta's Reality Labs, enhancing products like Ray-Ban smart glasses and supporting Meta's "personal superintelligence" vision. Limitless had raised about $35 million, and its CEO Dan Siroker expressed excitement about joining Meta. Following the acquisition, Meta stopped selling the pendant and ended the Rewind app, planning to integrate the technology with its Llama AI model.

Amazon boosts AI with new models and AWS edge

WIRED's Uncanny Valley podcast discussed Amazon's growing ambitions in artificial intelligence. Amazon, known for investing $8 billion in Anthropic, is also developing its own advanced AI models. The company has a strong advantage in the AI race because its AWS cloud service provides crucial computing technology. Amazon recently revealed new Large Language Models, including Nova Light, Nova Pro, Nova Sonic for real-time voice, and Nova Omni for experimental reasoning. They also introduced Nova Forge, a customizable LLM designed for specific business needs.

Poetiq AI system beats Gemini 3 Pro at lower cost

A small startup called Poetiq achieved impressive results on the ARC-AGI-2 AI benchmark, outperforming Gemini 3 Pro at half the cost. Poetiq's system reached 54% accuracy on December 5, 2025, by cleverly using existing large language models instead of building its own. The company developed a "meta-system" that finds the best ways to get information from these models through an iterative problem-solving process. This approach improved accuracy and lowered costs across 12 different models from major AI companies. Poetiq has open-sourced its code, showing that advanced AI might come from combining existing systems more effectively.

High school students debate AI in education

Knight High School students held their annual December debate, focusing on the pros and cons of artificial intelligence in education. Teacher Jane Thomas's English class discussed how AI tools like ChatGPT are changing learning. Students on the "pro" side argued AI can act as 24-hour tutors, support teachers, and help solve problems faster. The "con" side, which won the debate, raised concerns about misinformation, privacy, job losses, and AI's potential to weaken human creativity. They also noted the high cost of implementing good AI systems and its limited emotional intelligence.

Judge limits Google default contracts to one year

Judge Amit Mehta in Washington ruled that Google must limit its default contracts to one year. This decision supports the US Justice Department's efforts against Google's business practices. The ruling means Google can no longer pay companies like Apple and Samsung to exclusively set its search engine as the default on browsers and phones for longer than a year. Judge Mehta previously found that Google's contracts violated US antitrust law after a 10-week trial. Google plans to appeal the initial ruling, while the Justice Department might appeal this new remedy decision.

Canadian politician arrested for threats blames AI

Ontario Councilor Corinna Traill was arrested and charged with two counts of uttering threats. This follows a claim by former mayoral candidate Tom Dingwall that Traill left him a threatening voicemail in August, telling him not to run for mayor. Dingwall alleged the voicemail included threats of violence against him and his wife. Traill denied sending the message in a September Facebook post, stating she believed it was "created using artificial intelligence." She was released and is expected in court in January.

Investor warns AI FOMO causes bad decisions

Private equity investor Orlando Bravo warned that the excitement around artificial intelligence is causing "enormous" anxiety among investors. Speaking at the Bloomberg New Voices event in Miami on Friday, December 6, 2025, Bravo stated that a "fear of missing out" or FOMO is pushing private market firms to make hasty decisions. He believes this intense hype could lead companies to invest in AI tech firms they might later regret. Investors are struggling to properly evaluate which AI companies are truly worth the risk.

LinkedIn says AI not killing jobs despite warnings

Elon Musk and Bill Gates have both warned that artificial intelligence could eliminate many jobs within the next 20 years. However, a LinkedIn executive shared different findings from recent hiring data. LinkedIn's information shows that the demand for jobs is actually increasing, not decreasing. This trend suggests a more positive outlook for the job market, especially for younger generations like Gen Z.

Goodwill offers free Google AI training

Goodwill of the Finger Lakes, in partnership with Google.org, now offers free AI training to the community. The online Google AI Essentials course teaches people how to use generative AI in the workplace and takes about six hours to complete. This training is available at no cost until the end of 2026, and participants earn a certificate from Google. Jennifer Lake, President and CEO of Goodwill of the Finger Lakes, stated the program helps individuals adapt and grow in a changing workforce. Over 80% of past participants reported improved job performance after taking the course.

Sources

• Meta Snaps Up AI Wearables Startup Limitless
• Meta Acquires Limitless to Accelerate Work on AI-Enabled Wearables
• Meta Acquires AI Startup Limitless and Ends Hardware Sales | Ukraine news
• Meta Acquires AI Wearable Startup Limitless in Latest Hardware Push
• Meta Acquires Limitless to Advance AI Wearables and Smart Glasses
• WIRED Roundup: DOGE Isn’t Dead, Facebook Dating Is Real, and Amazon’s AI Ambitions
• Poetiq Goes Past Gemini 3 Pro On Arc-AGI At Half The Cost With Third-Party LLMs
• Artificial intelligence open for debate
• Google Must Limit Default Contracts to One Year, Judge Rules
• Canadian politician arrested after claiming threatening voicemail was AI
• Private Market FOMO Over AI Will Lead to Mistakes, Bravo Says
• Elon Musk and Bill Gates warn that AI will kill all jobs within 20 years. ‘That’s not what we’re seeing,’ LinkedIn exec says
• Free AI training now available to the community through Goodwill