PLUS: who’s really making money from AI and the US government’s new AI hiring spree
Good morning
Nvidia, long the dominant force in AI hardware, is making a significant push into the software ecosystem. The company just released its powerful Nemotron 3 family of models as open-source, signaling a major strategic play beyond just selling chips.
This move, combined with the acquisition of key infrastructure software provider SchedMD, shows a clear effort to build an entire ecosystem around its hardware. Will this two-pronged strategy of providing both models and tools succeed in making Nvidia's platform indispensable for developers?
In today’s Next in AI:
Nvidia's major push with Nemotron 3 models
How AI crawlers are reshaping the internet
The AI That Can't Be Deleted
The US government’s new AI hiring spree
Nvidia's Open Source Power Play
Next in AI: Nvidia is making a major move into AI software, releasing its powerful Nemotron 3 family of open-source models and acquiring SchedMD, a key provider of workload management tools.
Decoded:
Nvidia released its Nemotron 3 models in three sizes (30B, 100B, and 500B parameters) and is also providing the training data and tools to help developers build and customize AI agents.
The release positions Nvidia against other major players and popular Chinese open models, which now account for around a third of all tokens on some platforms, highlighting their growing importance.
To support this software push, Nvidia also acquired SchedMD, whose Slurm software is critical for managing massive AI computing jobs, a move that strengthens its role in the underlying infrastructure as the US allows exports of powerful chips to China.
Why It Matters: Nvidia is strategically moving beyond just selling chips to building an entire open-source ecosystem around its hardware. This two-pronged approach of providing powerful models and essential management tools aims to make its platform indispensable for the next wave of AI development.
AI Is Swallowing the Internet
Next in AI: Cloudflare's 2025 Year in Review shows AI is fundamentally reshaping the internet's landscape. Automated crawlers harvesting data for AI models now account for a massive and rapidly growing share of all web traffic.
Decoded:
AI-driven "user action" crawling, which mimics human behavior to gather data, exploded with a 15x increase in volume this year alone.
The scale is dominated by existing players, with Google's dual-purpose bot responsible for 4.5% of HTML requests—more than all other AI crawlers combined at 4.2%.
Website owners are pushing back, as AI crawlers have become the most frequently blocked agents in
robots.txtfiles, signaling a growing tension over data access.
Why It Matters: The insatiable appetite of AI for training data is placing new demands on global internet infrastructure. This trend is accelerating the shift toward a web where non-human traffic is the norm, creating new challenges for security and content monetization.
The AI That Can't Be Deleted
Next in AI: Microsoft's Copilot AI is now being automatically installed on LG smart TVs through a software update, and users are discovering there is no option to remove the new application.
Decoded:
This move is likely part of Microsoft's strategy to expand its AI footprint beyond PCs, potentially setting the stage for a wider rollout on other Linux-based systems.
The unremovable app adds to a growing trend of AI-driven data features on TVs, similar to LG's own "Live Plus" setting which uses viewing information for personalized ads.
The exact function of a powerful AI assistant on a TV remains unclear, highlighting a push for platform dominance over a clear, immediate user benefit.
Why It Matters: This move shows how aggressively tech giants are embedding AI into every device, whether a clear user benefit exists or not. Expect to see more ambient AI integrations as companies compete for a dominant position in the AI ecosystem.
Uncle Sam Wants You... for AI
Scott Kupor - director of the Office of Personnel Management
Next in AI: The US government is launching the "US Tech Force" to hire an initial cohort of 1,000 AI and tech experts. The two-year program aims to modernize public services and compete for top minds in the global tech race.
Decoded:
This initiative is a direct response to the intense war for AI talent, with the government offering competitive salaries between $130,000 and $195,000 to attract top minds.
Tech Force members will work on high-impact projects, like integrating AI into defense systems or building out the new Trump Accounts platform for the IRS.
This public-private partnership is a key part of the administration's broader AI action plan, with tech giants like Microsoft and Meta providing mentorship to cohort members.
Why It Matters: This move signals the government is getting serious about competing with the private sector for top tech talent. For professionals, it offers a unique chance to tackle massive, complex problems while building a powerful public and private sector network.
AI Pulse
Google drew criticism after its AI search summaries began generating dangerous and plagiarized recipes, causing significant traffic and revenue loss for food bloggers.
Mercor announced it reached $500M in annualized revenue, revealing the massive, lesser-known business of providing specialized human data for training advanced coding models.
SPSA expanded its partnership with AMP Robotics to create the country's largest AI-powered recycling project, aiming to divert half of the region's waste from landfills.