You’re receiving this email because you subscribed to the newsletter. If you’d like to unsubscribe, you can manage your preferences here with one click (or use the button at the bottom of the email).

Welcome to this edition of Loop!

To kick off your week, I’ve rounded-up the most important technology and AI updates that you should know about.

‏‏‎ ‎ HIGHLIGHTS ‏‏‎ ‎

Let's jump in!

1. Growing gap between CEOs and employees on AI productivity

We start this week with new research, which highlights a growing gap between CEOs and employees on AI adoption.

According to most employees, these tools have saved them less than 2 hours a week - or in some cases they’ve saved no time at all. Meanwhile, over 40% of executives claim that the technology is saving over 8 hours per week.

The gap also extends beyond time savings and into sentiment. Workers are far more likely to feel anxious or overwhelmed about AI than excited, while the C-suite reports the opposite.

This is unsurprising, given the rising cost of living around the globe and executives constantly talking about how they plan to eliminate jobs. No wonder that employees are feeling uneasy.

There’s also a real gap between rhetoric and reality. While the technology can be wonderful for some tasks, it can be really slow for others and force you to spend more time than you would on your own.

One UX engineer described spending an entire afternoon on what should have been a 30-minute AI-assisted coding fix, while another task that previously took days was completed in 20 minutes. This unpredictability makes it hard to identify where it’s improving productivity.

And finally, there’s the problem of “AI slop”. This is when employees use the technology to create basic, poor-quality work that needs to be fixed later on - often by other employees.

Unfortunately, this is a widespread problem that’s impacting every organisation. With any new technology, people will naturally try and use it to cut corners. That’s fine if the results are good, but if you copy-and-paste without fixing the mistakes - don’t be surprised if your colleagues get a bit annoyed.

We should use the technology to boost our work and explore problems from different angles, not outsource it completely - which is where this gap between CEO expectations and reality is.

2. Humans& raises $480 million to change how we work together

The new AI company, which was founded three months ago, has just raised a whopping $480 million in seed funding - with a $4.48 billion valuation.

It’s being backed with investment from Jeff Bezos, Nvidia, Google Ventures, and several other companies.

Their team includes top researchers from Anthropic, Google, xAI, and Stanford - with a focus on accelerating the work that teams can do, rather than eliminating jobs completely.

We don’t know much more about the company and the areas they plan to research, but it’s raised a shocking amount of money within just three months. They’ve clearly made good progress recently, if they’ve been able to convince investors to write such large cheques - even with that impressive lineup of researchers.

Much of that money will be spent on poaching staff from other AI labs and training their own models, with significantly more computing needed to create the most advanced AI models. It’ll be interesting to see what they release in the coming year.

3. Investors rush to fund “physical AI”

Continuing with the theme of huge investment, Ethernovia has just raised $90 million in Series B funding to expand its work on autonomous vehicles.

The company develops Ethernet-based processors that can quickly send sensor data to the central computer, which is critical since these cars need to react immediately and stop a car in milliseconds - rather than seconds.

But what’s really interesting are the investors backing it. Porsche, Qualcomm, and Maverick Capital have joined and this deal reflects a broader shift in where AI money is heading.

After years of pouring capital into foundation models and chatbots, investors are now chasing what's being called "physical AI" - the convergence of machine learning with robotics and autonomous systems.

That pivot is pushing funding downstream to suppliers and component makers who've traditionally operated well below the radar.

As the industry moves from impressive demos to actual deployable hardware, the companies that are solving mundane-sounding problems - like data latency and sensors - are attracting much greater levels of investment.

4. OpenAI responds to backlash over AI data centres

As local communities start to raise concerns about AI infrastructure and the environmental impact, OpenAI has agreed that it will cover the energy costs and minimise water usage.

I’m surprised this wasn’t their policy until now, as it’s common sense - companies should be paying for this, not ordinary taxpayers.

The company's statement arrives as local opposition to AI facilities intensifies, with residents increasingly frustrated by rising bills and the strain it’s putting on water resources.

Water consumption presents another thorny issue, with data centres being notorious for guzzling vast quantities of drinkable water.

The announcement follows a similar pledge from Microsoft, which suggests that Big Tech is finally waking up to these concerns - as dozens of new data centres are built across the US.

5. Vercel’s tool allows us to build Generative UI components

If your team’s building AI agents, this is an interesting tool that they should try. Vercel has just released “json-render”, which allows AI models to show interactive UI components - rather than text.

The approach builds on the familiar concept of server-driven UI, where interface structures are fetched as JSON from a backend. The twist here is that the AI can generate that structure on the fly.

Developers will provide the model with a list of available components and their properties, the AI produces a JSON response, and the client then displays the UI. This allows you to add buttons, visualise data, and validate the user’s response.

Of course, Vercel isn't alone in exploring this. Google has recently published their work on the A2UI protocol, which I covered in more detail here. It’s part of a broader trend called Generative UI, which is when we allow the AI more control and it decides which UI component should be shown.

Previously, developers had to create structured workflows and anticipate every action that the user wanted to make. Instead, this allows us much greater freedom and it’s really useful when you can’t prepare for every user journey.

As I’ve flagged before, we will see a lot more around Generative UI in the coming year - with AI labs developing their own standards for chatbots.

Apple is working on an AI wearable

According to insiders, Apple is developing an AI-powered device that can be worn on your clothes. They plan to create a thin, circular device that’s roughly the size of an AirTag - but slightly thicker.

It's said to pack two cameras - one standard and one wide-angle - alongside three microphones, a speaker, and a physical button.

The timing feels significant, coming just days after OpenAI's Chris Lehane hinted at Davos that his company plans to unveil its own AI hardware later this year - which could be a pair of earbuds.

Apple appears keen to keep pace, with reports suggesting the company is accelerating development to compete with OpenAI. A 2027 launch is being floated, although that timeline feels ambitious given how early the project is.

The elephant in the room, of course, is Humane AI. The company was founded by Apple veterans and launched a very similar device in 2024, but the battery life was incredibly poor and the AI pin was slow at answering questions.

Just a year later, the company was sold to HP.

While Apple has much deeper pockets than Humane and a more loyal customer base, it remains to be seen if consumers will buy another AI wearable like this.

🛰️ Jeff Bezos creates a Starlink rival, promises 6Tb uploads

📚 Google now offers free SAT practice exams, powered by Gemini

📺 Netflix earned $1.5 billion from ads in 2025

💻 Microsoft staff are using Claude Code to create software, not Microsoft Copilot

👶 Ex-Google staff are building an AI learning app for kids

📦 Zipline raises $600 million to expand drone delivery business

📄 Adobe Acrobat allows you to edit files with prompts

🏥 Serve Robotics acquires company that makes robots for the hospital

🚕 Tesla launches robotaxi rides in Austin, with no human safety driver

🚗 Waymo launches its robotaxi service in Miami

This defence startup is based in France and has raised over $200 million at a $1.4 billion valuation, despite being only one year old. President Macron has praised the announcement as "excellent news for France’s strategic autonomy”.

The company once dubbed itself "a European Anduril” that will challenge the defence incumbents, but it has pivoted recently and signed a partnership with Dassault.

Under these plans, the two French companies will work on next generation fighter jets and drones - with a focus on sovereign technology that’s isolated from the United States. It follows rifts within the NATO alliance and Europe is now moving to prioritise domestic manufacturers.

Sovereign AI, and technology more generally, will likely become huge growth areas for European companies in the future - but the UK and EU need to invest significantly more in their industries to become self-reliant.

With the ongoing war in Ukraine, drone attacks have served as a stark wake-up call for NATO countries - with startups working with traditional forces and helping them adapt more quickly.

Harmattan has already partnered with Ukrainian drone maker Skyeton, with plans to expand further into drone interception, electronic warfare, and ISR.

With fresh backing from Dassault and NATO contracts already secured, Harmattan looks very well-positioned for the coming years and is one to watch.

This Week’s Art

We’ve covered quite a bit this week, including:

If you found something interesting in this week’s edition, please feel free to share this newsletter with your colleagues.

Or if you’re interested in chatting with me about the above, simply reply to this email and I’ll get back to you.

Liam

Share with Others

If you found something interesting in this week’s edition, feel free to share this newsletter with your colleagues.

Liam McCormick is a Senior AI Engineer and works within Kainos' Innovation team. He identifies business value in emerging technologies, implements them, and then shares these insights with others.