AI: Apple Google 'Big AI' deal? RTZ #299
...and Apple MM1 SLM AI shapes Siri 2.0 'Small AI' moves
Appleās AI strategy outlines are coming into focus, piece by piece. We knew it was coming. Especially since Apple CEO Tim Cook had uncharacteristically signaled to expect Apple āAI Plansā soon on the companyās last earnings call. Bloombergās Mark Gurman today reports āApple Is in Talks to Let Google Gemini Power iPhone AI Featuresā:
āCompanies considering AI deal that would build on search pact. Apple also recently held discussions with OpenAI about deal.ā
āApple Inc. is in talks to build Googleās Gemini artificial intelligence engine into the iPhone, according to people familiar with the situation, setting the stage for a blockbuster agreement that would shake up the AI industry.ā
āThe two companies are in active negotiations to let Apple license Gemini, Googleās set of generative AI models, to power some new features coming to the iPhone software this year, said the people, who asked not to be identified because the deliberations are private. Apple also recently held discussions with OpenAI and has considered using its model.ā
This as and when finalized, is an eminently logical, and āwin-winā deal for both parties, especially as Apple already use Google as the default Search engine in its two billion plus active laptops, smartphones and other devices. And gets paid over $15 billion a year for the privilege. As Bloomberg continues to outline:
āIf a deal between Apple and Google comes to fruition, it would build upon the two companiesā search partnership. For years, Alphabet Inc.ās Google has paid Apple billions of dollars annually to make its search engine the default option in the Safari web browser on the iPhone and other devices. The two parties havenāt decided the terms or branding of an AI agreement or finalized how it would be implemented, the people said.ā
Apple of course seems to be still in negotiations around this arrangement ahead of its June WWDC Developer conference. And other Foundation LLM AI companies like OpenAI, Anthropic and others are likely also in the discussion mix. Again, from Bloomberg:
āWhile the talks between Apple and Google remain active, itās unlikely that any deal would be announced until June, when the iPhone maker plans to hold its annual Worldwide Developers Conference. Itās possible that the companies donāt reach an agreement or Apple ultimately chooses to go with another generative AI provider, like OpenAI. Or Apple could theoretically tap multiple partners, as it does with search in its web browser. Other generative AI providers include Anthropic, which offers a chatbot called Claude.ā
But Googleās Gemini family of LLM AI products make the most sense for Apple, not the least of the reasons being Googleās massive AI Google Cloud infrastructure around the world. This is critical to actually serve AI requests from Appleās over two billion users, especially on the āReinforced Learning Inference Loopsā that are critical to actually deliver meaningful and relevant AI results to users. Especially in a world where AI GPU infrastructure is currently scarce, and Nvidia is the global gatekeeper and allocation master via its founder/CEO Jensen Huang, for the forseeable future.
This deal with Google as and when finalized would address Appleās AI needs on the Foundation LLM āBig AIā side that Iāve been discussing for industry players. Apple of course remains in the pole position on āSmall AIā, with its own efforts on smaller, more efficient āSLMā (Small Language Models).
In that context, the company separately released research details of its upcoming āMM1ā SLM efforts this week as well:
āApple researchers have just revealed details of what could be the basis of this next-generation Siri, and if rumors are true, could work alongside Gemini on the iPhone offering a choice.ā
āReleased as a preprint research paper, MM1 essentially offers a new method for using AI-generated data and labels to speed up training of new models ā including possibly Siri 2.0.ā
āAt the core MM1 is a new method for training multimodal models using synthetic data including images and text.ā
āThe researchers behind MM1 claim their new method speeds up performance and reduces the number of follow up prompts to get a desired result.ā
āBeing able to improve prompt understanding and get to the desired output with as little interaction with the AI as possible is perfect for consumer tech, especially in Siri which will be used by a wide group of people with varying degrees of technological prowess.ā
āThe models achieve state-of-the-art pre-training metrics and competitive performance on multimodal benchmarks after fine-tuning.ā
āMM1 seems to be a family of AI models, with the largest around 30 billion parameters. This is significantly smaller than the trillion plus parameters in GPT-4 and Claude 3 Opus but the researchers still claim to match key benchmarks due to improvements in efficiency.ā
So Appleās AI moves seem poised to taking shape and take off.
With these and other AI related developments and platforms Iāve discussed before, Apple remains in an under-appreciated big beneficiary in these early days of the AI Tech Wave. Itās rapidly covering its options in both Big and Small AI. And Google also continues to be in a strong position to counter OpenAI, Microsoft and others in AI augmented Search, as Iāve discussed at length before.
This year continues to be proceeding at a breakneck pace in AI developments. Later today we get more details on Nvidiaās AI GPU and infrastructure roadmap as well. Lot more to come on AI moves by players big and small. Stay tuned.
(NOTE: The discussions here are for information purposes only, and not meant as investment advice at any time. Thanks for joining us here)