AI Top-of-Mind for Dec 7
For those following the state of play, a big announcement yesterday with Google unveiling Gemini, their new multimodal models. Some performance numbers below, as published in the Google DeepMind paper. For clarity, MMLU and MMMU are two recent AI benchmarks. And a detailed analysis by Alberto Romero in ‘The Algorithmic Bridge’ as well as ‘NY Times’ coverage. But doesn’t it seem that the FAAMGs are biting their teeth even further into our brains?
Not to be overshadowed, AMD also announced yesterday their AI chipset roadmap. ‘CNET’ released a short video, while ‘Tom’s Hardware’ reviewed the event in detail. A few snippets:
· Generative AI will require significant investments to meet the needs for training and inference workloads. One year ago, AMD predicted a $150 billion TAM for AI workloads by 2027. Now AMD has revised that estimate up to $400 billion in 2027.
· The MI300 delivers performance parity in training with Nvidia but exhibits the strongest advantages in inference. AMD highlights a 1.6X advantage in inferencing.
· Norrod invited representatives from Arista, Broadcom, and Cisco to stage to talk about the importance of continued adoption of the Ethernet standard for data centers.
· If you’re wondering why this is important — Nvidia has acquired Mellanox and uses its Fibre Channel networking gear heavily in its systems. Notably, Nvidia is not a member of the Ultra Ethernet consortium.
Turning to creative, I had the opportunity to test out Leonardo.Ai’s ‘Live Canvas’ feature that attempts to create imagery based on your doodles or photo uploads. There is a ‘creativity’ setting where you adjust the value from .3 to .9, and needless to say, the results were a bit interesting.
And what type of regulation can we expect and what challenges await? The NY Times published a good analysis, and two salient paragraphs:
Even now, E.U. lawmakers are arguing over what to do, putting the law at risk. “We will always be lagging behind the speed of technology,” said Svenja Hahn, a member of the European Parliament who was involved in writing the A.I. law.
The absence of rules has left a vacuum. Google, Meta, Microsoft and OpenAI, which makes ChatGPT, have been left to police themselves as they race to create and profit from advanced A.I. systems. Many companies, preferring nonbinding codes of conduct that provide latitude to speed up development, are lobbying to soften proposed regulations and pitting governments against one another.
Finally, from the marketing front, ‘Ad Age’ discusses advantages of digital twins for market testing and personalization. And, how this could lead to our ‘avatars’ making routine decisions for us.
