- AI KATANA
- Posts
- Meta unveils biggest Llama 3 AI model
Meta unveils biggest Llama 3 AI model
Also: A Sam Altman-Backed Group Studied Universal Basic Income For 3 Years. Here’s What They Found.

Morning!
In today’s newsletter, Meta unveils its largest AI model yet, the Llama 3, featuring 405 billion parameters and multilingual capabilities, aiming to surpass competitors by next year. Generative AI is set to significantly impact remote white-collar workers, altering job landscapes post-pandemic. A study on universal basic income reveals positive shifts in recipients’ focus on meaningful work and stability. Despite advancements in AI chatbots, they still struggle with math, highlighting ongoing AI development challenges. AI self-regulation shows progress but remains inconsistent. A survey indicates the majority of recent graduates feel unprepared for the workforce due to a lack of AI training in college. In venture capital news, Canadian AI startup Cohere secures $500 million in funding to expand its enterprise-grade AI technology.
Sliced:
🆕 Meta unveils biggest Llama 3 AI model
👩🏻💻 Gen AI is Coming for Remote Workers First
💵 A Sam Altman-Backed Group Studied Universal Basic Income For 3 Years. Here’s What They Found.
➗ A.I. Can Write Poetry, but It Struggles With Math
🤖 How’s AI self-regulation going?
👩🏻🎓 Majority of Grads Wish They’d Been Taught AI in College
Meta has launched its largest Llama 3 AI model, featuring 405 billion parameters and multilingual capabilities. This new version surpasses its predecessor in language proficiency, computer code generation, and solving complex math problems. Despite being smaller than some competitor models, such as OpenAI’s GPT-4, Meta aims for Llama 3 to overtake proprietary models by next year. Alongside the flagship model, updated versions of the 8 billion and 70 billion parameter models were released, all featuring expanded context windows for better handling of extensive user requests. Meta’s strategy of offering these models for free aims to spur innovation, reduce dependency on competitors, and boost engagement on its social platforms, though some investors are concerned about the associated costs. Early benchmarks indicate Llama 3 is competitive with leading models from Anthropic and OpenAI, and Meta plans to release multimodal versions incorporating image, video, and speech capabilities later this year.
The advent of generative AI is set to impact remote white-collar workers significantly, a shift from the historical trend where automation primarily affected blue-collar jobs. The rise of remote work during the Covid-19 pandemic granted white-collar workers more autonomy, but now generative AI is poised to change their job landscape. This technology can perform tasks such as drafting emails, creating content, and analyzing data, potentially reducing the need for human input in these areas. As generative AI becomes more integrated into remote work environments, white-collar workers may face job displacement and a shift towards more automated and less varied roles, mirroring the automation trends seen previously in blue-collar sectors.
A three-year study funded by Sam Altman, CEO of OpenAI, has explored the effects of universal basic income (UBI) by providing $1,000 monthly payments to participants across Illinois and Texas. The findings revealed that recipients showed a modest decrease in employment but increased their focus on meaningful work, education, and job training. The study also noted improvements in healthcare spending and housing stability, with participants reporting higher expenditures on basic needs like food and transportation. Despite a slight drop in employment hours, the UBI group exhibited a greater interest in job searching and moving to better neighborhoods. This research, backed by $14 million from Altman and $10 million from OpenAI, suggests that UBI could play a significant role in addressing the economic shifts brought about by advancements in AI. Future findings will delve into the impacts on politics, relationships, household composition, and children’s education.
AI chatbots like OpenAI’s ChatGPT have showcased impressive abilities in language tasks but continue to struggle with math, a discrepancy highlighted by their design focus on probabilistic language generation rather than strict rule-based calculations. This divergence from traditional computing, which excelled in precise mathematical tasks, underscores the fundamental shift brought about by neural networks that learn from data rather than following programmed rules. Efforts to address these math challenges include integrating calculator programs, as seen with Khan Academy’s Khanmigo and ChatGPT, to enhance accuracy. The debate within the AI community about the path to achieving artificial general intelligence (AGI) highlights differing views on the reliance on large language models versus broader system approaches. Despite their limitations, AI chatbots remain valuable in various applications, demonstrating the evolving landscape of AI technology and its ongoing development.
A year after the White House secured voluntary AI commitments from top tech companies, the landscape of AI self-regulation shows both progress and limitations. While eight more companies have signed on and efforts such as AI red-teaming, watermarking, and safety research have been implemented, the nonbinding nature of these commitments means compliance can be inconsistent. The US remains reliant on tech companies’ goodwill to prevent AI harms, as legislative action on AI regulation is unlikely in the polarized political environment. The future direction of US AI policy hinges on the upcoming presidential election, with contrasting approaches from candidates potentially shaping the regulatory landscape. Despite these challenges, the voluntary commitments have helped establish norms and pressure companies to prioritize responsible AI development.
A new survey by Cengage Group reveals that 70 percent of recent college graduates believe generative AI should be included in college curricula, with over half feeling unprepared for the workforce due to a lack of training in AI tools. The survey of 974 graduates, including nontraditional learners, highlights a generational disparity in AI preparedness, with 61 percent of Gen Z graduates feeling unprepared compared to 48 percent of millennials and 50-60 percent of older generations. Additionally, 39 percent of graduates feel threatened by the potential of AI to replace their jobs. Employers echo these concerns, with 62 percent expecting hires to possess foundational AI skills, and more than half indicating a preference for candidates with AI experience. Despite some integration of AI in university systems, faculty adoption remains slow, with only 14 percent confident in using AI in teaching. The study suggests that cross-sector collaboration and updated curricula are essential to bridge this skills gap and better prepare students for an AI-driven job market.
🧑🏽💻 AI Jobs
🛠️ AI tools updates
The Tom’s Guide Awards 2024 celebrated the year’s best AI tools and devices, recognizing significant advancements in generative AI. Claude was named the best AI chatbot for its robust features, including advanced reasoning, coding abilities, and innovative Artifacts feature. Leonardo stood out as the best AI image generator for its photorealistic image creation and unique Phoenix model. Runway earned the title of best AI video generator, praised for its realistic motion and video quality in the new Gen-3 model. Udio was highlighted as the best AI music generator for its superior sound quality and customization options. The Meta Ray-Ban Smart Glasses were awarded the best AI device for their practical AI functionalities, including real-time translations and impressive photo and video capabilities. Lastly, Circle to Search on Android was recognized as the best AI phone feature for its seamless integration and convenience in identifying objects and translating text.
💵 Venture Capital updates
Canadian AI startup Cohere has secured $500 million in a Series D funding round led by PSP Investments, valuing the company at $5.5 billion. The funding, which saw participation from new investors such as Fujitsu, Cisco Systems, EDC, and AMD Ventures, will be used to expand teams and enhance Cohere’s enterprise-grade AI technology. Known for its emphasis on data security, Cohere develops cloud-agnostic AI solutions tailored for various industries, offering flexible deployment options. Founded in 2019, the company operates from Toronto and San Francisco, with significant offices in London and New York. Cohere’s client base spans sectors including banking, technology, and retail, and its revenue surged from $13 million in 2023 to $35 million in March 2024. The company plans to double its 250-employee workforce within the year.
🫡 Meme of the day

⭐️ Generative AI image of the day

Before you go, check out AI is confusing — here’s your cheat sheet.
