POLICY

Trump's EO nationalizing AI lacks teeth

Let the lawsuits begin. President Trump signed his promised executive order seeking to limit states’ ability to regulate AI. 

While the president likely lacks the legal ability to preempt state lawmaking, the order will establish an AI Litigation Task Force, restrict federal funding and grants for states that pass “onerous AI laws,” and attempt to preempt state lawmaking through powers granted to the Federal Trade Commission. Already, state leaders are threatening legal action should the Trump administration enforce the order.

Senator Ted Cruz took up the cause of limiting states’ intervention in AI over the summer, but the moratorium was kiboshed. Now, Trump is taking the executive order route after lobbying from the tech industry and White House AI and Crypto Czar David Sacks, who claimed that a patchwork of state AI regulations would handicap the US in the global race for AI dominance. 

State-by-state regulation can be onerous for businesses. In the auto industry, for instance, carmakers have historically had to grapple with emissions standards that vary from one state to another within the US. But critics of the action say that Team Trump’s crackdown on state AI legislation fails to address safety concerns for Americans and children in particular.

“With this Executive Order, Trump is delivering exactly what his billionaire benefactors demanded — all at the expense of our kids, our communities, our workers, and our planet,” Massachusetts Senator Ed Markey said. “A broad, bipartisan coalition in Congress has rejected the AI moratorium again and again, and I intend to keep that streak going.”

While there have been over 1,000 AI regulation bills proposed at the state level, per the WSJ, it’s still hard to tell what exactly the effects of strict state-level AI regulation might be. Nevertheless, it looks like lobbying against state regulation may already be paying off. New York Governor Kathy Hochul plans to strike the text of the recently passed RAISE Act and replace it with SB 53, a more lenient AI law passed in California, Transformer reported.

For those favoring a “one rulebook approach” to AI regulation, this probably won’t move the needle. Executive orders lack the teeth of congressional law. Plus, as Semafor’s Ben Smith pointed out on TBPN, Trump’s recent embrace of the AI industry with little public messaging could embolden anti-AI Republicans in the midterms.

TOGETHER WITH CEREBRAS

20× Faster Inference, Built to Scale

Advanced reasoning, agentic, long‑context, and multimodal workloads are driving a surge in inference demand—with more tokens per task and tighter latency budgets—yet GPU‑based inference is memory‑bandwidth bound, streaming weights from off‑chip HBM for each token and producing multi‑second to minutes-long delays that erode user engagement.

Cerebras Inference shatters this bottleneck through its revolutionary wafer-sized chip architecture, which uses exponentially faster memory that is closer to compute, delivering frontier‑model outputs at interactive speed.

Start building what others can’t.

BIG TECH

Nvidia + HPE’s AI alternative to AWS and Microsoft

Nvidia and Hewlett Packard Enterprise (HPE) are sharpening their pitch that AI factories are the next big unlock for enterprise AI.

At the AI Summit in New York City, representatives from both companies argued that businesses need to evolve beyond traditional data centers, which were built around training large language models, and toward AI factories that are engineered for the power-hungry inference workloads behind applications. With the right architecture, they said, organizations can inject advanced AI into everyday workflows across HR, legal, R&D, and operations with fewer bottlenecks than they face in legacy compute environments.

“Traditional compute architectures aren't working,” Matthew Hitt, OEM Sales Manager at Nvidia, said during the December 11 panel. He pointed to network chokepoints and power and cooling limits that are slowing down adoption. “The AI factory has a new thought process behind it, of ‘power in, tokens out.’”

Their pitch comes as Nvidia and HPE embark on a charm offensive to expand their joint ecosystem to build out AI factories. In December, they opened their first AI factory lab in Grenoble, France to serve European customers who want local control over their infrastructure and data rather than relying on large US cloud providers. The move marks one of the first major buildouts since the companies established their AI focused partnership in June 2024.

Onstage, Hitt and Christopher Smith, a sales leader rep for HPE’s cloud, services, and AI division, highlighted GPUs that sit at the core of these deployments. That includes the RTX Pro Server for inference workloads, HGX B300 for training, and its GB200 NVL72 rack of GPUs. They positioned these as building blocks for both private-cloud AI environments and larger factory-scale builds.

The vision still comes with constraints. AI factories draw significant energy and water and require specialized engineering that can slow the path from pilot to production. The speakers pointed to HPE’s custom architecture work, sustainability dashboards, and liquid cooling as attempts to smooth those barriers.

With the concept still early, Hitt said enterprises will need partners to make these systems real. “You're not going to be successful on your own,” he said. “This is too complex.”

AI factories also dovetail with the growing sovereign cloud movement, as governments and enterprises in Canada and the EU push to keep sensitive data and compute within national and regional borders. By packaging high-performance AI infrastructure into localized, privately operated environments, Nvidia and HPE are offering a potential alternative for some workloads to US giants like AWS and Microsoft that dominate the global cloud market.

TOGETHER WITH FLORA

This Platform Has Every Creative AI Tool, All In One Place

It’s called FLORA, and it isn’t your standard creative platform. FLORA is like the Photoshop to every other platform’s MS Paint – sure, they’re both technically for design, but one offers a laundry list of best-in-class tools trusted by pros everywhere, while the others… well, the others can be kinda fun to goof around in.

Designed for professionals who demand control, speed, and precision (all without sacrificing creative control), FLORA’s seamless workflow enables iteration and collaboration at scale. It’s the first AI tool built by creatives, for creatives – and now, you have a chance to try it.

The first 50 people to use code DEEPVIEW at checkout will get 50% off their first month of FLORA's Agency Standard Tier (monthly option only), so get moving right here.

AI TRAINING

Turn a text prompt into a full song in 60 seconds with Suno

With Suno, you can create a royalty-free, original song with vocals—ready to use in YouTube videos, podcasts, ads, or TikTok—without touching an instrument or DAW. We’ve covered a lot on Suno’s lawsuits, funding and partnerships, so to start off this new section of the newsletter, we thought it’d be great to show you what Suno does!

How to do it:

1. Go to suno.com, sign up free, and click Create. Toggle to Custom mode at the top—this unlocks lyrics and structure control.

2. In Style of Music, layer: genre + mood + instruments + vocals. Be specific—the more detail, the better the output.

3. Add lyrics with metatags to structure your song. Key tags: [Intro] [Verse] [Chorus] [Build] [Drop] [Outro]

4. Pro subscribers: Open Advanced Options and set Weirdness to 20-30% for clean output, Style Influence to 70%+ to lock in your sound. (Pro is $10/mo or $8/mo yearly—free users can skip this step)

5. Hit Create—Suno generates 2 versions in ~30 seconds. Regenerate 4-6 times until you nail it, then download the MP3 from the menu.

CULTURE

Leaders tell how to future-proof jobs from AI

The AI boom will not wipe out all jobs.

At the AI Summit in New York City, leaders across HR, edtech, and social media shared frank views on which roles will endure and what skills workers need to stay competitive as automation spreads.

Since January of 2025, tech companies have eliminated nearly 50,000 U.S. jobs, and named AI as a key driver, according to a November study from HR consulting firm Challenger, Gray & Christmas. Mariano Allegra, SVP of commercial strategy at tech recruitment firm Lawrence Harvey, said that while AI may be “eating away” at the job market, he’s seeing a hiring surge in defense and robotics.

Allegra said those clients are “kicking up” compensation to pull talent from other industries. Professionals who understand AI are especially sought after, including “AI trainers” who can embed the technology into corporate workflows to boost productivity.

Some workers are already using AI to expand their opportunities. Yesim Saydan, an independent AI and communication strategist, said she was able to take on more clients after she began creating custom chatbots. Working with entrepreneurs to build their LinkedIn brands, Saydan fine-tunes each GPT to match their voice and business goals so it can draft content in their preferred style, freeing up time for strategy.

Saydan said the rise of agentic AI will make the ability to manage these “AI employees” a core workplace skill. “People with AI skills who know how to manage agents and make different agents work together is a skill that will be needed,” she told The Deep View.

During the panel, Daniele Grassi, president and CEO of General Assembly, an upskilling platform, argued that the greatest risk comes when companies fail to reskill the talent they already have. He pointed to a financial services client that kept its talent by retraining employees whose roles were cut and placing them in UX design and product management.

“Lifelong learning,” Grassi said, is the only reliable way for workers to stay competitive as AI reshapes the future of work. 

Behind the optimism around reskilling, AI-driven job replacement is becoming an increasingly attractive option for companies looking to do more with fewer workers. A 2025 survey from AiResumeBuilder.com found that roughly 3 in 10 firms expect to replace workers with AI in 2026, with more than half seeking to automate at least 10% of their roles across teams such as IT and accounting. For employees whose jobs are at risk of automation, knowing how to leverage AI to boost productivity may be necessary to keep their jobs.

LINKS

• OpenAI has reportedly eliminated the equity vesting period for new employees

• The AI data center boom is pulling resources away from other critical construction projects

• Taiwan opens AI data center with supercomputer powered by Nvidia Blackwell chips

• El Salvador partners with xAI to bring Grok to public schools

• Amazon Prime cuts AI-powered recaps of shows after users point out inaccuracies 

• Bond markets are flashing warning signals over massive AI data center buildout

• Legal startup Solve Intelligence raises $40M Series B to build tools for IP and patent law work

• Dex: Organizes your messy browser tabs, remembers your progress and task actions for you.

• Alli Studio: A video reasoning model powered by Sora 2, Runway and Veo 3 for creating studio-grade visuals.

• Tnkr: Collaboration platform where robotics builders can document, share, and contribute to open-source robot projects—from hardware to AI models.

• Google Vids: AI-powered tool that turns your ideas and files into polished work videos in minutes.

• OpenAI: Research Engineer, Retrieval & Search, Applied Engineering

• Snowflake: AI Research Scientist - Reinforcement Learning

GAMES

POLL RESULTS

Is physical AI the next ChatGPT?

• Yes, it's about to take off (25%)

• Not yet, still too early (42%)

• No, different kind of problem (33%)