Meta’s AI edge
M1 releases AI Marketing Assistant
Meta’s AI edge
Meta will soon get a 100K GPU supercomputer to train Lama 4
For Those Who Seek Unbiased News.
Be informed with 1440! Join 3.5 million readers who enjoy our daily, factual news updates. We compile insights from over 100 sources, offering a comprehensive look at politics, global events, business, and culture in just 5 minutes. Free from bias and political spin, get your news straight.
***
M1 releases AI Marketing Assistant
AI Marketing Assistant, covers the full cycle of tasks: from creating an Ideal Customer Profile, building a marketing strategy, to creating custom ads and posts for social media.
All this in the required tone of voice and taking into account the marketing goals of the business.
Users highly rate Ideal Customer Profiles, which allow them to better understand the audience, its goals, problems, etc. AI Marketing Assistant also does research to provide about 20 places where the selected target audience spends time.
On average, users note an increase in key marketing metrics (CTR, open rate, conversion) from 48%.
You can try AI Marketing Assistant for free
***
Meta’s AI edge, to incorporate improved AI into its Ray-Ban smart glasses
Meta's(Facebook’s) Ray-Ban smart glasses have emerged as a sleeper hit, thanks largely to their audio and video capabilities rather than their existing AI features. However, CEO Mark Zuckerberg recently showcased upcoming AI upgrades that could offer more useful services for regular people than most current AI chatbots.
For instance, a Ray-Ban wearer will be able to talk to someone speaking a foreign language and get a real-time translation through the glasses' speakers, albeit with a slight delay. Perhaps even more impressive, users can look at the number of a parking spot they've just parked in and ask the AI assistant to remember the number - a simple feature that could win over many consumers.
Neither Google nor OpenAI has a product that can directly compete with Meta's smart glasses. While having apps available on a smartphone is not the same, Zuckerberg has previously boasted about the AI in the Ray-Bans, despite it being substandard experience. The company is now hoping the new AI upgrades will work as well as the recent demonstration suggested.
However, there are concerns about the cost of these AI features, which Meta is offering as a free add-on. Chief Product Officer Chris Cox claimed the company's new Llama 3.2 large language model is cheaper for developers to use than any other, implying low operating costs. But at mass scale, those costs could still add up significantly….
Read the original on Techradar
***
Meta will soon get a 100K GPU supercomputer to train Lama 4
Meta Platforms is on the verge of completing a massive supercomputer cluster that will significantly boost its AI capabilities, particularly for training the next iteration of its large language model, Llama 4.
The new supercomputer cluster is reported to consist of slightly more than 100,000 GPUs, specifically NVIDIA H100 server chips. This massive array of high-performance GPUs will provide Meta with unprecedented computational power for AI training and development.
The primary purpose of this supercomputer cluster is to train Llama 4, the next version of Meta's Llama large language model. This aligns with Mark Zuckerberg's earlier statements about Llama 4 requiring ten times more computational power than its predecessor, Llama 3.2.
The cost of the GPU chips alone is estimated to exceed $2 billion. This is part of Meta's larger AI investment strategy, with capital expenditures projected to reach up to $40 billion this year. The supercomputer cluster is in its final stages of completion. It is expected to be fully operational by October or November 2024. The cluster is located somewhere in the United States.
Meta aims to position Llama 4 as the industry's most advanced model by next year. The AI infrastructure is designed to support various Facebook AI needs also including content ranking and recommendations. While these investments may not yield immediate revenue, they are expected to pay off in the long run….