Everything We Missed This May

Our top stories this month.

We were gone for a bit, but the tech world didn't wait for us, thankfully. Layoffs kept coming with big AI spending right behind, Anthropic took the lead in AI race as OpenAI ramps up efforts, Japan being Japan with robots, and to no one's surprise, lots more.

Here are our top picks from the past couple of weeks. Better late than never.

Big Layoffs vs Big AI Spend

Over 100,000 tech workers lost their jobs in the last few months. LinkedIn cut roughly 5% (~875 jobs), Cisco cut 4,000, PayPal cut 4,800, and Meta is cutting another 8,000 soon. And it's not just these companies, there's more to come across the board. At the same time, Amazon, Microsoft, Google, and Meta are expected to collectively spend roughly $725B+ on AI infrastructure and data centers in 2026 depending on whose projections you use. Big number. Huge.

But the big question: are companies firing staff to pay for compute? The infrastructure race is dizzying to watch, but it's not that simple. AI tools are taking a major cut of the blame for this one. There's the junior dev thing too. With the aggressive hiring the early 20s saw, maybe we're still looking at the fallout. Maybe.

One thing's certain, AI has a death grip on the payroll budget.

The AI (Two-Horse?) Race

At the start of the AI race, OpenAI had a 1–2 year headstart over most companies around the GPT-3/GPT-3.5 era. Before ChatGPT, most AI companies were publishing papers, building niche tools, etc. but OpenAI turned a frontier model into a mainstream consumer product overnight.

They deserve their chops for that.

But the gap started shrinking fast once Google activated Gemini (aggressively too), Anthropic scaled Claude, open-source models exploded, and GPU access spread. Last year saw OpenAI with 32% of US business adoption. Anthropic had 8%. That gap is now gone. Ramp's May 2026 AI Index shows Anthropic at 34.4% of business adoption, crossing OpenAI at 32.3% for the first time ever. How? Two words.

Claude Code.

Anthropic's coding tool reached $2.5 billion in annualized revenue by February 2026 alone. Together, OpenAI and Anthropic now control 89% of generative AI startup market revenue. OpenAI still has distribution and brand recognition on its side, so it's not like they're going down without a fight. We'll see how this plays out.

Microsoft Antitrust Pressure

Word on UK streets: Microsoft allegedly charges businesses significantly more to run their standard software (like Windows Server) on a competitor's cloud (like Amazon AWS or Google Cloud) than it does on its own cloud (Microsoft Azure). The result? A pending $2.8 billion lawsuit.

Here's the whole story. Britain's Competition and Markets Authority (CMA) launched an investigation into Microsoft's dominance in the business software space. They're saying customers are locked-in (not the way you think) because Microsoft practices allegedly reduce competition in cloud services by putting rivals at a disadvantage. Microsoft bundles its dominant products (Windows, Word, Excel, Teams) with new services like its AI assistant (Copilot) so if a company wants to keep using the industry-standard Excel or Outlook, they have to accept the terms.

If you're a company in that situation and a rival AI can't plug into your Office data, you're less likely to switch away from Microsoft's own Copilot. This is what's under scrutiny.

AI is changing the market. The CMA is launching this investigation now because it's probably easier to intervene before a monopoly, rather than fining a company after all's said and done.

Africa's Digital Infrastructure Push

Silicon Valley spent the last few weeks fighting over AI models and market share, but another race is being run in the background. Across Africa, governments, telecom companies, and cloud providers have been pouring money into data centers, fiber networks, fintech rails, and AI-ready infrastructure. The digital hype isn't just about startups anymore, it's becoming about power, compute, and connectivity.

In Kenya, Microsoft and G42's planned $1 billion AI data center project hit pause because powering the facility could require "switching off half the country." It's not cancelled, but AI infrastructure needs huge amounts of electricity, and many countries are still building the grids required to support it.

Nigeria is also pushing heavily into cloud and AI infra. Lagos and Abuja are seeing multiple large-scale data center projects underway as the country positions itself as a West African compute hub. Plus major fintech and digital investment, Lagos is reinforcing its role as one of Africa's biggest tech ecosystems.

Africa's tech story right now isn't just apps anymore. Hasn't been for a while now. It's electricity, it's compute too, and it's also who gets enough infrastructure in place before the AI era fully settles in.

TikTok's Global Infrastructure Expansion

TikTok. Expansion. Data centers. Three continents.

Thailand: TikTok is reportedly spending around $25 billion USD here for expanding server, data storage, and processing facilities across three provinces. Yes, Bangkok is one of the three. Why? The country has over 50 million TikTok users (about 75% of its 66-71 million population). TikTok Shop Thailand GMV reached $52 billion in Q1 2026. We too would build data centers there if we were seeing those numbers. This is the largest single overseas data center investment by any Chinese internet company to date.

Finland: Under Project Clover, TikTok is investing another $1.17 billion investment to build a second data center. The first one is expected online by the end of 2026.

Why? Cold climate = less cooling costs.

Brazil: This will be the first data center in the Americas, and it's still in progress. Also, with an estimated 131 million users in Brazil, ByteDance has requested the go-ahead from Brazil's Central Bank to operate as a credit and payments fintech. They're seeking two licenses: a payment institution license (for digital accounts and transfers) and a direct credit company license (for issuing loans).

The company is transitioning from lightweight *lease what you need* operations to heavyweight *build your own* infrastructure across multiple continents at once.

The Expected AI Job Boom

AI Translators. Do they speak fluent AI? Sort of. This is one of the few roles companies suddenly can't hire fast enough. What do they do, you ask? They connect AI tools with existing business systems: banking software, manufacturing platforms, healthcare systems, or enterprise workflows. So, amid the tech job bloodbath, these guys are sort of sitting in a puddle.

AI still makes mistakes. Companies now want engineers who can supervise AI agents, catch hallucinations and bugs, communicate with teams, and turn AI output into usable products. In a lot of cases, one very skilled engineer with AI tools can now handle work that previously required entire teams.

The takeaway? Technical skills alone are no longer enough. Speak the language of your replacements or pack it up.