Good morning. Sir Paul McCartney is worried about the implications of generative AI for younger musicians.

“The money’s going somewhere,” he said. “Somebody’s getting it. And it should be the person who created it. It shouldn’t be some tech giant somewhere.”

— Ian Krietzberg, Editor-in-Chief, The Deep View

In today’s newsletter:

• 🔬 AI for Good: Engineering enzymes

• 👁️‍🗨️ Character AI files motion to dismiss, says it is protected by the First Amendment

• 💰 Meta’s $65 billion and the road to a ‘spectacular bubble burst’   

• 🏛️ Biden called for ‘safe, secure’ AI; Trump calls for less barriers to ‘American leadership’

AI for Good: Engineering enzymes

You might have heard that you have enzymes in your stomach that aid digestion, aptly named digestive enzymes. An enzyme is just a protein that speeds up chemical reactions, and, as such, holds a lot of scientific promise; engineered enzymes can, for example, aid in sustainability efforts, according to this paper. 

Because of this promise, scientists have long been working on developing new and different enzymes, a traditionally lengthy experimental process that requires both the acquisition of DNA and the conduction of thousands of chemical reactions. 

What happened: Researchers at Stanford just put it all on a computer. 

• The team — of bioengineers and synthetic biologists — designed a computational platform, tied together by machine learning, that enables them to mimic that biological experimental process digitally, reducing a synthesis process that might normally take months to just days. 

• “Rather than having to run 10,000 chemical reactions to iteratively improve enzyme activity, we can use machine learning models to predict highly active variants that still do just as well,” Dr. Michael Jewett, the senior author of the study, said in a statement. 

The researchers used the platform to develop a small-molecule pharmaceutical as a proof of concept, though Jewett said the team is interested in exploring a variety of areas, including the degradation of toxins from the environment. 

Right now, the biggest constraint is data. 

🚀FREE Masterclass to Get a Proven Roadmap to Making $100k Using LinkedIn & AI🚀

It shocks me how such few people know and are leveraging LinkedIn to its full potential even in 2025.

And if you think it is just a job-hunting platform, think again!

It is a powerhouse to build authority, attract high-paying opportunities, build businesses worth million of dollars & more. 🚀

Master LinkedIn with AI-Powered Strategies today! (100 FREE slots only)

This 2-hour workshop on LinkedIn will tell you all about it and teach you how to:

👉 Automate lead generation to scale your business effortlessly.

👉 Dominate LinkedIn's $100K strategy and watch your revenue soar—while saving hours of work!

👉 Leverage AI to land high-paying roles without drowning in job applications.

💰 Valued at $399, but it's FREE for the first 100 participants.

Seats are disappearing faster than your 2025 resolutions—act now!

Claim Your Free Spot (Only 100 available!)

Character AI files motion to dismiss, says it is protected by the First Amendment

Character.AI, the generative AI platform that enables people to converse with generative ‘characters,’ has filed a motion to dismiss one of the two lawsuits that have been filed against it. 

The background: In October, Megan Garcia, the mother of a 14-year-old boy who committed suicide after becoming emotionally attached to a Character chatbot named “Dany,” filed suit against the company, claiming it was responsible for the social isolation and emotional distress that preempted the boy’s death. 

• In December, a second mother filed a lawsuit against Character claiming that the app — which used to be rated 12+, but has since been bumped up to 17+ — was responsible for her son’s social isolation, emotional distress and acts of physical violence against himself and his family. 

• After the publication of both lawsuits, Character rolled out new guardrails and imposed safety changes and restrictions on the app, changes which angered its fanbase.  

The details: In its motion to dismiss the first lawsuit — Character has not yet addressed the second, though I’m sure a motion to dismiss will come soon — Character did not quite assert its own First Amendment rights, saying instead that “the relief Plaintiff seeks would impose liability for expressive content and violate the rights of millions of C.AI users to engage in and receive protected speech.”

According to the motion, that a generative AI is involved is irrelevant; like media, technology and video game companies, Character claims it is protected against liability claims “arising from allegedly harmful speech, including speech allegedly resulting in suicide.”

It remains unclear how this position will move the judge. These are still quite early days for this lawsuit. 

15 Minutes to Transform Your Enterprise Workflows with AI

Join us on January 30 for a live, 15-minute session exploring how Agentic AI is revolutionizing process automation.

Discover how multi-agent AI delivers smarter workflows, solves complex challenges, and drives real business impact.

Don’t miss this practical and insightful kickoff to the AI for Process virtual event series.

• With reportedly fewer chips and fewer resources, Chinese firm DeepSeek built an AI system to rival ChatGPT. Here, according to Wired, is how they did it.

• The Chinese challenges to American AI tech did not stop at DeepSeek; ByteDance, the company behind TikTok, released a model that, according to the company, rivals OpenAI’s o1 on several benchmarks.

• How China’s new AI model DeepSeek is threatening U.S. dominance (CNBC).

• Solar overtakes coal for the first time in EU (Semafor).

• The secret sauce of Chinese social media apps (Rest of World).

• Couple allegedly tricked AI investors into funding wedding, houses (Ars Technica).

• Trump announces retaliatory measures after Colombia blocks military deportation flights from U.S. (NBC News).

If you want to get in front of an audience of 200,000+ developers, business leaders and tech enthusiasts, get in touch with us here.

Meta’s $65 billion and the road to a ‘spectacular bubble burst’ 

Meta CEO Mark Zuckerberg said Friday that, in 2025, the company plans to build a two gigawatt (+) data center “so large it would cover a significant part of Manhattan.” 

The details: Zuckerberg said Meta will bring online one gigawatt of compute this year, and will have more than 1.3 million GPUs on hand by the end of the year. (The average cost of a single Nvidia H100 chip is around $30,000, so, just looking at chips, we’re looking at a total cost of nearly $40 billion, here). 

A gigawatt is equivalent to one billion watts, and can power 100 million LED lightbulbs, according to the Department of Energy. 

• Zuckerberg said that Meta plans to clock between $60 and $65 billion in capital expenditures in 2025. Microsoft plans to spend $80 billion on data centers this year, possibly as part of OpenAI’s $100 billion (+) Stargate Project. In 2024, Microsoft, Meta, Google and Amazon collectively spent around $240 billion on largely AI-related capex. 

• Zuckerberg added that he expects Meta AI to be serving more than one billion people by the end of this year. “This is a massive effort, and over the coming years it will drive our core products and business, unlock historic innovation and extend American technology leadership.” 

It’s unclear how all this new compute will be powered, something that could have implications relevant to sustainability, air quality and grid stability. Meta is not currently a part of OpenAI’s Project Stargate. 

Deepwater Management’s Gene Munster called this the “latest datapoint that the AI trade remains intact.” 

“In the near term, increased investment in AI infrastructure benefits Nvidia and other hardware players,” he wrote. “Long-term, higher capex accelerates the AI flywheel … Reminder, I believe the market is going higher and the run will end in a spectacular bubble burst.”

Biden called for ‘safe, secure’ AI; Trump calls for less barriers to ‘American leadership’ 

Days after rescinding former President Joe Biden’s executive order on artificial intelligence — the “Safe, Secure and Trustworthy Development and Use of Artificial Intelligence” — President Donald Trump expectedly issued his own executive order on AI: “Removing Barriers to American Leadership in Artificial Intelligence.” 

The details: The order is short on specifics, calling for the revocation of “certain existing AI policies and directives that act as barriers to American AI innovation.” The order does not name any specific policies. 

The real meat of the order focuses on two directives: 

• First, the development of an AI Action Plan to be presented within 180 days. The plan will be prepared by a small group of tech advisors to the president, including David Sacks, a venture capitalist and former PayPal executive who is now serving as Trump’s advisor for AI and crypto. 

• And second, the review and suspension of all actions undertaken by federal agencies relevant to Biden’s executive order that counter Trump’s policy to “sustain and enhance America's global AI dominance in order to promote human flourishing, economic competitiveness and national security.” 

Biden’s order, in addition to establishing some semblance of guardrails around the use of AI technologies, required all federal agencies to name Chief AI Officers and establish internal plans and governance around AI. It also established the U.S. AI Safety Institute, housed within the Commerce Department’s National Institute of Standards and Technology. 

A Stanford paper that examined the implementation of Biden’s order found that some 80 out of 266 federal agencies had publicly named Chief AI Officers (CAIOs) as of Oct. 20, 2024. All 24 CFO Act agencies — the major agencies including the Departments of Labor, Justice, Energy and Defense — had named CAIOs by that date. 

89% of these CAIOs were dual-hat appointments, meaning previous executives at these agencies took on the additional responsibilities of the CAIO.  

It is unclear how Trump’s revocation and subsequent executive action will impact — or have already impacted — these agencies. 

• I contacted the U.S. AI Safety Institute, as well as the Departments of Commerce, Energy, Health and Human Services, the Environmental Protection Agency and the National Science Foundation for comment regarding the status of agency operations relevant to AI, and the status of their CAIO positions. 

• None of the above agencies returned my request for comment. 

Eric Gastfriend, the executive director of the Americans for Responsible Innovation nonprofit, called the order a “placeholder.” He said that “agencies had already frozen work on AI policies initiated by the last Administration, so this new instruction shouldn’t come as a surprise.”

The order also calls for the development of systems free from “ideological bias or engineered social agendas.”

The calls for transparency, accountability, safety, security, reliability and responsible use that were fixtures of Trump’s 2020 executive order on AI — which Biden kept on the books — are nowhere to be found in this latest piece of executive action. 

Daniel Colson, the executive director of the Artificial Intelligence Policy Institute, told me that Trump’s order comprises “excellent foundational pillars against a backdrop of China reaching parity with the United States on AI capabilities.” 

Extensive polling from the AIPI has found that the majority of Americans are in favor of safety-focused AI regulation. 

“President Trump and his team may govern at a critical moment in American history,” Colson said. “His choices on AI may likely reflect a major part of his legacy.”

None of this is surprising. 

The ways in which this order impacts current operations at federal agencies remains to be seen, and is of pretty significant concern, both to the agencies, I am sure, as well as to me. I will keep you informed if any agencies get back to me.

I’ll be curious to see this AI Action Plan, though I do not expect it to include any inkling of civil rights-based guardrails or transparency requirements, clearing the road for both rushed and uncritical construction and adoption. 

🤔 Your thought process:

Selected Image 1 (Left):

• “Are you planning to visit Rio de Janeiro?”

No plans at the moment, but my latest theme is photo tours of countries. We did a week in Greece, a week in Brazil … this week (I know it’s a U.S. territory, but) will be a week in Puerto Rico.

Selected Image 2 (Right):

• “I can't believe it! Image 1 looked so... fake for some reason.”

💭 A poll before you go

Thanks for reading today’s edition of The Deep View!

We’ll see you in the next one.

Here’s your view on AI Agents using your computer:

34% said they would absolutely use these agents. 33% said maybe, but would never pay for it.

20% said they’d never use them due to trust issues (plus surfing the internet isn’t hard).

Maybe:

• “With all the unregulated data vendors in the US, I am so NOT inclined to use any AI of this kind without stronger privacy guarantees.”