Good morning. I do not think I have experienced what you might describe as a ‘slow news day’ in months. Anyway.

Yesterday marked the beginning of a global trade war, one perhaps accentuated by Ontario’s decision to rip up a $68 million contract with Elon Musk’s Starlink.

There will be many victims in this war. How AI will fare in this environment is a question that we don’t yet have an answer to. We do, however, have a few possibilities on the table.

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

In today’s newsletter:

🎙️ Podcast: The case for AI guardrails

On the latest episode of The Deep View: Conversations, I connected with Liran Hason, a machine learning engineer who founded an AI guardrails startup back in 2019 called Aporia.

Since I first connected with Liran in 2023, he’s been making the case for guardrails as a means of enabling — rather than limiting — artificial intelligence.

We get into the details: what guardrails are, how they work and why we need them.

Check it out below (or above):

AI for Good: Species identification 

Google has open-sourced SpeciesNet, an AI model designed to identify wildlife captured on camera. 

The details: The system has been available to conservationists through the website Wildlife Insights since 2019. The open-source release of the underlying model, Google said, will allow researchers to scale up their monitoring efforts. 

The release coincides with Google’s launch of a startup accelerator for companies building solutions to protect the environment, a push that clashes with Google’s massive energy, water and carbon footprint, a footprint that has only grown due to its investments in AI. 

Why it matters: Conservation decision-making, from public policy approaches to specific actions undertaken by understaffed, underfunded researchers, needs to be informed by detailed environmental realities. Understanding the movements of species within changing habitats arms researchers with the knowledge needed to better inform their conservation efforts.

Is AI Actually Making Test Automation Faster? The Data Might Surprise You

AI is everywhere in software testing—but is it actually speeding things up? 

Rainforest QA surveyed 625 developers, engineers, and tech leaders to uncover what’s working (and what’s not) in test automation.

In their report, you’ll learn:

Unlock the data-backed insights that can help you streamline testing and deliver faster. Don’t wait—download the full report today!

Waymo partners up with Uber in next step of steady expansion

The news: Uber users in Austin, Texas, can now be automatically matched up with one of Waymo’s self-driving robotaxis. 

Uber said that users will have the option of denying a proffered Waymo robotaxi if they’d prefer a human driver. 

At this point, the Uber Waymos can cover 37 square miles of Austin, though Uber said there are plans to expand this operating region. Uber said that the next Waymo-Uber partnership will be hitting Atlanta, Georgia, though didn’t specify a date. 

The landscape: This latest expansion follows a massive year for Waymo, during which the robotaxi company proved that it doesn’t have any real competition. The self-driving company — whose cars are powered by a stack of technology including radar, Lidar, cameras and artificial intelligence — delivered four million rides in 2024, after delivering only 700,000 in 2023. 

Its rollout has been relatively cautious, though self-driving experts have warned that, as Waymo expands, it remains unlikely that its safety record will expand in kind. 

As for Uber, the company tried once to get self-driving off the ground, but shuttered the program after one of its cars was involved in a fatal accident.

What the trade war might mean for AI 

The tariffs that President Donald Trump has been hyping up since the early days of the campaign have now officially gone into effect. 

(A tariff is essentially a tax on imported goods, one that often leads to higher inflation and higher costs for consumers). 

What’s the damage: For starters, Trump imposed a 25% tariff on all goods imported from Canada and Mexico, the country’s top two international trade partners.

Mexican President Claudia Sheinbaum said the country will respond with tariff and non-tariff measures to be announced on Sunday. 

Trump has vowed to impose additional tariffs as soon as April 2. He has discussed targeting the European Union with tariffs similar to those put in place against Canada and Mexico. 

The stock market was in full retreat Tuesday — the Dow fell 1.5%, the S&P fell 1.2% and the Nasdaq, after staging a big comeback, closed down .3%, marking a second consecutive day of stock market losses.

What this means for AI: Even as this policy measure comes into focus, there are such a variety of factors at play that it is truly impossible to predict the full impact this will have on the tech sector, AI companies, AI development and the AI bubble. 

Wedbush analyst Dan Ives said in a note Tuesday that this will be an “uncomfortable time for growth investors.” 

For tech companies, the real challenge is the supply chain; tariffs could significantly drive up the cost of manufacturing, especially for chipmakers reliant on Chinese exports. And that could drive prices up even more.

But the Oxford economists don’t expect the tariffs to stay in place for long, a point that Ives echoed, saying that the tariffs are an attempt to bring “countries to the negotiating table.” 

The problem with this assumption is that it’s not clear what Trump is negotiating for; he’s justified the tariffs by pointing to the apparent influx of fentanyl into the U.S., a flood that, at least in Canada’s case, is really more of a bare trickle (the U.S. seized 43 pounds of fentanyl at its northern border in 2024, and the majority of the 21,148 pounds of fentanyl seized at the Mexican border was taken from American citizens). 

Still, the impact here is one that will be accentuated, or lessened, depending on how long these tariffs remain in place, something that we have little clarity on. Case in point, as of last night, Commerce Secretary Howard Lutnick hinted that we might have tariff compromises in effect as soon as today. But again, we just don’t know what’s going to happen, or when.

Ives believes that neither tariffs nor policy adjustments change the scope of the “AI Revolution” or the scale of the capital expenditures hyperscalers will pour into it. Tariffs, he said, could “slow spots down in the near-term, but the winners of this AI Revolution … will remain the stocks to own and any weakness is a buying opportunity.”

Nvidia rallied a bit Tuesday, rising about 1.6% to regain just a bit of lost ground. Shares of Microsoft, though largely flat, hit a 52-week low Tuesday, and shares of Tesla sank a further 5% to fall a total of 32% for 2025 so far.

Depending on the amount of pain being felt economically, we might see a reduction in AI startup investment, and we might likewise see a reduction in enterprise purchases of the technology, something that can hurt the AI ecosystem. In a highly inflationary environment, money becomes tight and spending patterns tend to become a bit more cautious; enterprises have been struggling to find the monetary value associated with expensive GenAI tech for two years, and many have failed.

In an inflationary environment, interest in it might well recede.

What is unlikely to change is Big Tech’s capex approach, which is supported by a variety of established money-winning businesses, a strong insulation against a weakening economy. That said, tech stocks tend to do poorly during times of heightened inflation, and investors, already nervous about high valuations and low AI-related revenue, don’t seem like they’ll be on board with tens of billions in AI infrastructure expenses if it starts to hurt the bottom line.

Played out for long enough, and coinciding with government cuts that have hit the National Science Foundation, among other things, we might see a big unraveling of the AI trade, that circular ecosystem that is built on demand for chips, demand that could quickly begin to falter.

However, at least on the AI side, I think much of this won’t really impact consumers.

The great shield that the developers have here is that, tariffs or not, AI is a massively money-losing gambit. 

Whether you charge $20 per month or $200 per month, whether your GPUs cost $30,000 or $40,000 per chip, the scale of the investment required for AI isn’t being recouped; more expensive chatbots, for instance, will just push people away, when the developers are currently trying hard to do the opposite.  

This environment is not a good one for the development of AI. That much is clear. What remains unclear is the ways (and speed) in which things — startup investment, capex, valuation come-downs, infrastructure build-outs, enterprise demand, etc. — will change.

🤔 Your thought process:

Selected Image 1 (Left):

Selected Image 2 (Right):

💭 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 Apple and GenAI:

A third of you really want a GenAI-enabled Siri and are confident that Apple will get it right in the end. A third of you don’t care about a GenAI-enabled Siri, and so don’t mind the slower approach, here.

17% of you really want a GenAI-enabled Siri and are confident that Apple is fumbling it.

Something else:

If you want to get in front of an audience of 450,000+ developers, business leaders and tech enthusiasts, get in touch with us here.