Google Pays Samsung, Adobe Integrates OpenAI & Google, Baidu's New LLM, & AI Education Executive Order

Google is paying Samsung "substantially" for Gemini installs, Adobe's Firefly integrates OpenAI & Google, Baidu launched their upgraded LLM, and a new executive order for AI education

Google Pays Samsung for Gemini AI App Installs

Google’s antitrust trial revealed that they have been paying Samsung substantial monthly sums to preinstall Gemini AI on Galaxy devices, as revealed during a U.S. antitrust trial.

The deal, potentially extending to 2028, ensures Gemini’s prominence on Samsung phones, offering users features like real-time translation and task automation.

Google’s Android operating system is open, but their AI is vertically integrated. We recently wrote about how they’re the only leading AI provider who is not reliant on Nvidia’s GPUs, using their own TPUs instead.

Distribution will be critical for a vertically integrated provider. Just like how Apple found AT&T as a partner when releasing their iPhone, Google is signing up distribution partners as well.

Adobe Integrates OpenAI and Google AI into Firefly App

Adobe announced that its Firefly app integrates OpenAI’s ChatGPT and Google’s Imagen 3 alongside their own AI models.

Firefly is Adobe’s suite of generative AI models and tools, where users can generate images, videos, and design assets from text prompts and simple inputs

These integrations allow users to generate images and videos using multiple AI systems, enhancing creative flexibility for designers and marketers.

They’re taking a platform approach, offering users multiple options for enhancing images. Given that Adobe isn’t an AI industry leader, this makes much more sense than being vertically integrated like Google.

The next question is whether AI only competitors such as OpenAI or Midjourney can erode Adobe’s pricing premium, or if there will always be a premium for Adobe’s Creative Cloud.

Baidu Launches Ernie 4.5 Turbo AI Model

Baidu released their Ernie 4.5 Turbo model, intensifying competition in China’s AI market.

The model excels in natural language processing and multimodal tasks like image analysis, targeting enterprise applications in e-commerce and healthcare. It offers cost-effective performance for developers, but their focus on Chinese-language tasks could limit global reach.

Trump Signs Executive Order on AI Education

President Trump signed an executive action to bolster AI education and workforce training, underscoring AI’s role as a national priority.

The order directs federal agencies to expand AI-related curricula in schools and fund retraining programs for workers impacted by automation. It aims to address the growing demand for AI skills in sectors like healthcare and logistics.

While the initiative has broad support, critics highlight the absence of clear funding commitments, which could hinder timely rollout.

Another challenge is that we don’t know where AI is going. If you asked any expert during the dot com boom who would win the internet, only the few who guessed Amazon would have been right. Even in the early 2000s, that company’s market cap was around $4 billion, so it was far from obvious.

As for Google, they received their A round from Sequoia in 1999, and Facebook was founded in 2004.

The internet was a new paradigm, and we know that most of the winners didn’t exist when the boom started. AI is the new paradigm now, and our policy makers should similarly not assume they know who or what the winners will be.

xAI Enhances Grok 3 with DeepSearch and Voice Mode

xAI announced upgrades to Grok 3, including vision, multilingual audio, and realtime search in voice mode.

These features position Grok 3 as a versatile tool for researchers and consumers, but is it enough to capture users from OpenAI?