Hello!

This week's top story looks at Oracle, the largest tech company you might not have heard of. This week they laid off 30,000 employees across their global workforce to lean hard(er) into AI. We'll take a little "medium dive" into why you should care.

Also this week, we take a look at Meta (again), the White House's new app, and quantum navigation.

And for our blog spotlight, a reminder of what technology should be like, in the form of the Fediverse.

Let's dig in...

Top Story

🖥️ What's going on at Oracle...

A quick summary

After careful consideration of Oracle's current business needs, we have made the decision to eliminate your role as part of a broader organizational change. As a result, today is your last working day.

That was the email that greeted 30,000 Oracle employees on Monday morning this week. In departments across the US, Canada, India, and other regions, individuals arrived into work and were immediately locked out of key systems. It's estimated that this will ultimately free up between $8-10 billion from the company's cash flow.

Who are Oracle anyway?

You'd certainly be forgiven for not recognising the name. They aren't part of the so-called "Magnificent Seven" (Apple, Amazon, Alphabet (Google), Meta (Facebook), Microsoft, NVIDA, and Tesla), but most of them will have Oracle systems in their stack.

You, in fact, will almost certainly have come into contact with Oracle's systems today. Their bread and butter is in corporate database systems. From banks to hospitals, train operators to supermarkets, Oracle provides part of the infrastructure for a huge number of large companies. More recently, they've tried to leverage that pedigree and infrastructure to become a leading player in cloud computing. Which brings us back to the need for a little more cash than routing around the back of the sofa might provide.

This is not the first time Oracle has made cuts

Founded in 1977, Oracle has been around a minute. So this isn't the first time it's made layoffs. But it does seem to be an accelerating trend. This latest, and largest, round follows redundancies across the US and EMEA in 2023 (focused mainly on sales and support), and another round in 2024 (cloud engineering and, again, support).

It's worth noting that Oracle aren't the only ones making cuts. 2025 was a bad year for middle-income tech employees from the big firms. An estimated 153,000 jobs were cut across Microsoft, Intel, Amazon, HP and others in November alone. And the pace doesn't seem to have slowed into this year. Across the same rogues gallery of companies, by early March another 30,000 jobs had gone (most of them from Amazon), before the Oracle cuts this week.

So why all the cuts?

You can have three guesses, but you probably won't need them...

AI is the single biggest factor in these job cuts. Either the jobs themselves can be done by AI, or the companies need the money to prop up their insane AI spending. Most of the time, its both. But for Oracle in particular it's deeper than just making more money. Well... no it isn't. But what I mean is, its not just about freeing up some spare cash to feed the meter of the current system.

Larry Ellison, Founder, Chairman, CTO, and former CEO, of Oracle, recently went into a little more detail about his thoughts on the future of AI. His position is that we're on the plateau of what AI can do at the moment, because it is trained only on "publicly available information" - i.e. the internet. We'll pretend for the moment that Mr Ellison is not familiar with the countless lawsuits against companies like OpenAI, Anthropic, Microsoft and others, that would disagree with this, and focus instead on his wider point. Ellison thinks that in order to unlock the next level of AI, it needs access to private data.

Speaking on a Q2 earnings call, Ellison said:

The Oracle AI Data Platform makes all your data — ALL of your data — accessible to AI models. Not just the data in Oracle databases and Oracle applications, but data from other databases, cloud storage from any cloud, even data from your own custom applications are accessible to AI models using the Oracle AI Data Platform. Using our AI Data Platform, you can unify all your data and reason on all of your data using the very latest AI models.

And at the World Governments Summit in February, he said:

We have to move it all [a county's data] into a single, unified data platform. That’s the missing link—we need to unify all national data and put it in a database that is easily consumed by the AI model

In short, Ellison wants access to all of your information. Only then, he believes, will we get close to that holy grail of tech bros - Artificial General Intelligence.

Does this feel familiar...?

Once again, our top story of the week is about a company looking to bring disparate data sources together in order to provide "better information". (I promise I don't plan this...)

As as with last week's Palantir story, we're talking about a similarly broad spectrum of datapoints: finance, healthcare, transport, government, etc. I'm not suggesting that Oracle is stealing your data to put into one giant repository of information that they can use for their own ends. I'm just saying that if any company could do that, it would be easiest for them. And they wouldn't have to go through all the rigmarole of pesky contract negotiations with various government bodies. They're already in.

And look, to be clear, Oracle can't just go hoovering up data - especially not in the UK and EU, where we're protected by our respective GDPR laws. Even if a government body deploys Oracle, as the Data Controller they (the government body) is responsible for making sure that Oracle only uses that data for the stated purposes. Which I'm sure is of great comfort to you. And, as we know, the law is a fast and agile machine that has no trouble keeping up with changes in technology.

Also, and it feels late in the segment to be throwing out completely new information, but I'll do it anyway, Oracle are one of the key partners in "Stargate AI project" a $500 billion joint initiative between OpenAI, SoftBank, and Oracle to build a networked US-wide slew of AI data centres, all to develop the next generation of frontier-AI.

Larry Ellison isn't just spit balling about the future of AI, he's trying to build it. And if there's one thing he's not short of, it's the data to do it.

What else is happening in the world of tech?

💰 Landmark losses for Meta and YouTube as big tech misses the point

In what could arguably have been this week's top story, Meta (owner of Facebook, Instagram, Threads, and WhatsApp) have been fined nearly $400 million as it was found that they are liable for products that inflict harm on young people - including sexual exploitation.

In a separate case the next day, both Meta and YouTube were fined $6 million as both were found to have deliberately designed products that were addictive and meant to hook young users.

A jury in New Mexico ordered Meta to pay $375m in damages over claims that its products led to child sexual exploitation, among other harms. The following day, a jury in California ordered Meta and YouTube to pay $6m over claims that both companies deliberately designed addictive products to hook young users.

These two trails were the first in what might soon seem like an unending onslaught for big tech, with cases lined up against Meta, YouTube, TikTok and Snap.

While Meta and YouTube will both look to appeal the court decisions, the fact remains that this is the first time in pretty much the existence of algorithmic media that they've come out of court with their tail between their legs.

💻 I decompiled the White House's new app

The official White House Android app has a cookie/paywall bypass injector, tracks your GPS every 4.5 minutes, and loads JavaScript from some guy's GitHub Pages.

The White House have a new app, and it is everything you would expect from a neo-proto-fascist government. There's a lot of code snippets in this article. But even if you aren't in the mood to pretend you're in the Matrix, skip over those bits and keep reading - the content in between is very insightful.

Perhaps the most disturbing part of this whole thing is the location tracking. The app checks your location every four and a half minutes while you're using it, and every nine and a half when the app is in the background. In both cases, it sends that data to the app's servers, storing it alongside metadata such as which notifications you react to (and how), details about your device, and more. What it does with all of that amalgamated information is anyone's guess. But as readers of last week's newsletter will know, there are plenty of uses for it.

The app also pulls in JavaScript from an individual developer's GitHub repo. In case some of those words don't mean anything to you (and it is a nerdy sentence, even for me): one of the core parts of the app is loaded from a publicly accessible repository. In the grand scheme of things, that might not seem like the most important thing to write home about. However, if anyone gains access to this individual's GitHub account they could feasibly replace the code there with malicious code, which would (presumably) be loaded straight into the app without the need for approval.

🛰️ World-first quantum navigation tested on UK train to replace GPS tracking systems

The technology is being developed as an alternative to fixed trackside positioning infrastructure. Such systems can be expensive to install and maintain. It is also vulnerable to environmental disruption or equipment failure.

A new kind of navigation system is being tested on public trains for the first time. Our current method of tracking involves a mixture of expensive trackside equipment, and GPS.

It's not a stretch to understand why the government finds this important. With low-Earth orbit getting more and more cluttered (in large part due to the arms race between Starlink and competitors to sweep the market), and GPS being both poor at smaller distances and subject to interference from external powers, the main application of quantum navigation isn't going to be the 9.16 into Waterloo.

Sidebar: did you know that in 2025 we launched 4,510 objects into space, of which 82% came from US agencies and companies - including Starlink, Amazon, and others. This was the highest number on record, beating 2023's 2,903 objects by quite a margin.

Blog spotlight

👐 Openness, transparency and reach: three reasons why public institutions should embrace the Fediverse

It's been a big month for open source over on the continent. A consortium of EU-based partners just launched their open source office suite, and the German government mandated the use of open source file formats for public administrations.

In that spirit, my blog of the week comes from Elena Rossini, an Italian filmmaker and FOSS (free and open source software) enthusiast. She talks about why the Fediverse should be the default social media platform for public bodies.

It's a really clear and well structured argument that acts as a great reminder that good technology should serve you well. It should give you what you need and then get out of your way. Or, perhaps more accurately, it should never be in your way in the first place.

Bluesky hasn’t introduced ads yet - but it may eventually, in order to become profitable. [...] Its new interim CEO Toni Schneider shared in a blog post that "We are very actively thinking about how we will make money, but ads are near the bottom of the idea list." And yet I am reminded of Open AI CEO Sam Altman who changed his stance on ads. In 2024 he declared: "I think of ads as a last resort for us for a business model"; fast-forward to January 2026 and he is currently testing ads in ChatGPT. Netflix followed a similar trajectory.

Ads are just one example Rossini gives, and its a powerful one. Its not that the people designing the Fediverse are "better" than those running BlueSky and Meta (although...). Platforms like Mastodon are structured in such a way that it is inherently difficult to push things onto people that they don't want. It is so easy to move - and take all your "stuff" with you - that if there's anything you don't like about one platform you can just switch to another. It would be like moving all of your Facebook posts, followers, likes and comments, over to Instagram at the click of a button. You can't do that, and those two platforms are owned by the same company - that's how much they want to lock you in.

And if you read that and liked it, you might also appreciate Mat Duggan's Boy was I wrong about the Fediverse, and in case you missed it from a couple of weeks ago, Technically Good's Caught in a Multi-Billion Dollar Web: Why Leaving the Legacy Algorithmic Media is So Hard.

That's it for another week of LT3! Hope you enjoy a nice and sunny long weekend, full of chocolate if that's your thing.

I'll see you back here next week.

Will