IMO Weekly Highlights - 06172024

IMO picked up a few pieces of notable news about AI and tech companies for the past week. This week, Nvidia?has once again solidified its position as the undisputed leader in AI innovation with the release of “Nemotron-4 340B,” a groundbreaking family of open models that is set to revolutionize the generation of synthetic data for training large language models (LLMs).?

?1)??? Nvidia’s ‘Nemotron-4 340B’ model redefines synthetic data generation, rivals GPT-?

Nvidia?has once again solidified its position as the undisputed leader in AI innovation with the release of “Nemotron-4 340B,” a groundbreaking family of open models that is set to revolutionize the generation of synthetic data for training large language models (LLMs). This development marks a significant milestone in the AI industry, as it empowers businesses across various sectors to create powerful, domain-specific LLMs without the need for extensive and costly real-world datasets.

The model, which had been operating under the mysterious alias “june-chatbot” on?LMSys.org Chatbot Arena, has now been officially identified and introduced, stirring considerable buzz in the AI community.?

2)?? Microsoft delays broad release of Recall AI feature due to security concerns

On June 13th, Microsoft announced that it is?delaying the broad release?of its Recall artificial intelligence feature for Copilot+ PCs. Recall, which was originally slated to be widely available to Copilot+ PC users on June 18, will now first be released as a preview to members of the?Windows Insider Program?in the coming weeks.

The decision to push back the general availability of Recall stems from Microsoft’s desire to gather additional feedback and ensure the feature meets the company’s stringent security and quality standards before rolling it out to all users. The move underscores the growing scrutiny and caution surrounding the deployment of AI capabilities, as companies grapple with balancing innovation and responsible stewardship of the technology.

?3)?? New Transformer architecture could enable powerful LLMs without GPUs

Matrix multiplications (MatMul) are the most computationally expensive operations in large language models (LLM) using the?Transformer architecture. As LLMs scale to larger sizes, the cost of MatMul grows significantly, increasing memory usage and latency during training and inference.?

Now, researchers at the?University of California, Santa Cruz,?Soochow University?and?University of California, Davis?have developed a?novel architecture?that completely eliminates matrix multiplications from language models while maintaining strong performance at large scales.?

4)?? What we know about Apple’s on-device AI?

Following Microsoft Build and Google I/O,?Apple?was under a lot of pressure to show its on-device AI might at its Worldwide Developers Conference 2024. And as far as the demos are concerned, Apple has done a great job of?integrating generative AI?into the user experience across all its devices.

One of the most impressive aspects of the demonstrations was how much of the workload is taking place on the devices themselves. Apple has been able to leverage its state-of-the-art processors as well as a slew of open research to provide high-quality, low-latency AI capabilities on its phones and computers. Here is what we know about Apple’s on-device AI.?

5)?? ‘We don’t need Sora anymore’: Luma’s new AI video generator Dream Machine slammed with traffic after debut?

The fast-moving AI video generation market has shifted, yet again: Luma AI, a startup backed by famed Silicon Valley venture firm Andreessen Horowitz,?announced?the free public beta of its new AI video generation model,?Dream Machine, and it’s already faced a crush of users.

Though the model promises generations of up to 120 frames in 120 seconds (2 minutes, or a frame per second), the reality is that many users have been waiting hours in a digital queue on the Luma Dream Machine website for their video to process. This is due to sheer volume of traffic, according to the company.