🚨Turns out rolling back DEI did nothing🚨 

TL;DR: A new UC Berkeley study finds that S&P 500 companies which kept their DEI commitments performed as well as or better than those that rolled them back. So after all the “this is a financial necessity” talk, it turns out some companies may simply have folded under political pressure… for nothing. 💸

What Happened

Researchers compared companies that maintained DEI commitments  including Apple, Cisco, Costco, Delta, JPMorgan, Microsoft and Pfizer  with firms that scaled them back, including Citigroup, IBM, Target and Walmart. Looking at revenue and stock-market performance, the researchers found no financial penalty for keeping DEI, with some measures showing slightly stronger returns. 📈

The report’s conclusion is fairly awkward: “Large corporations appear to have folded under pressure for no financial gain.” In other words, if the argument was “we don’t necessarily want to roll this back, but shareholders demand it”… apparently the shareholders forgot to send that memo. 😬

That doesn’t mean every DEI programme automatically improves profits, nor does the study prove causation. But it does puncture one of the more convenient narratives of the past two years: that abandoning DEI was simply hard-headed business pragmatism rather than, in many cases, a response to political and legal pressure. 🧐

Why This Matters

The corporate retreat has been very real. A separate Catalyst and NYU study found 51% of federal contractors had reduced inclusion efforts, while 52% of companies without federal contracts were actually increasing them. Funny how quickly supposedly deep-rooted commercial convictions can change depending on who controls the procurement budget. 🏛️

And yet eight in ten US companies still say they remain committed to DEI. Many are changing the language, structure or legal framing rather than abandoning inclusion completely. So reports of DEI’s death may, once again, have been slightly exaggerated. ⚰️

The Bigger Picture

There’s a wider lesson here about corporate values. When DEI was fashionable, companies told us inclusion was fundamental to innovation, talent, performance and their values. When the politics changed, some suddenly discovered it was actually quite optional. 🔄

That matters particularly to employees from underrepresented backgrounds because these programmes were never just corporate branding exercises to them. They affected recruitment, promotion, sponsorship, pay equity, accessibility and who actually gets heard inside organisations. Removing them because the political weather changed doesn’t make those underlying inequalities disappear. 🌧️

And perhaps most awkwardly, the companies that stuck with DEI appear not to have been punished financially for doing so. So the choice may not have been values versus shareholder returns after all. It may simply have been values versus discomfort. 👀

🧠 So What?

Companies are obviously entitled to review ineffective DEI programmes. They should measure what works, cut performative nonsense and focus on outcomes. But that's very different from claiming that inclusion itself became commercially unaffordable. 📊

If anything, this research suggests something rather less flattering: some businesses panicked, changed course and got absolutely no financial reward for it. Congratulations on successfully abandoning your stated principles at zero ROI. 🏆

The bigger question now is what happens when the political winds change again. If your commitment to inclusion disappears whenever it becomes inconvenient, perhaps it wasn’t a strategy in the first place. It was marketing. 🎭

📚 Read More

HR Dive — Companies that roll back DEI don’t perform better financially, study says
https://www.hrdive.com/news/companies-dei-roll-back-dei-performance-uc-berkeley/828590/

UC Berkeley — Markets Do Not Punish Firms for Maintaining DEI
https://democracypolicylab.berkeley.edu/wp-content/uploads/2026/08/folsz_grumbach_dei.pdf

😢Grad jobs down😢

TL;DR: Graduate vacancies listed on Adzuna have fallen 46% in a year, hitting their lowest level since its records began in 2016. AI isn't the only culprit, but when entry-level work disappears, we should probably start asking how people are supposed to become experienced workers. 🪜

What Happened

Adzuna recorded just 8,383 graduate vacancies in July, down from 15,397 a year earlier. For context, its listings peaked at more than 55,800 graduate roles in 2017. 📉

Employers point to a fairly brutal cocktail: AI replacing or reducing junior work, higher employer National Insurance costs and minimum-wage increases making entry-level hiring more expensive. Competition is rising too, with 2.14 jobseekers for every vacancy across the market. 🤖

There is an important caveat: graduate-market experts say Adzuna's numbers may partly reflect companies no longer advertising roles specifically as “graduate jobs”. So no, the UK's graduate labour market hasn't necessarily shrunk 46% overnight. But even the experts agree: it isn't in good shape. 🫠

Why This Matters

Here's the slightly obvious problem with eliminating entry-level jobs because AI can do entry-level work:

How does anyone become senior? 🤔

Companies want experienced hires. But experience traditionally comes from being hired when you don't have much experience, doing some slightly boring junior work, learning how an organisation functions, making mistakes and gradually becoming useful.

Remove the bottom rung of the ladder and eventually you discover you've also removed the route to the top. 🪜

And Britain already has a youth employment problem. More than one million young people are now not in education, employment or training, while unemployment among 16-to-24-year-olds stood at 16.2% in the three months to March. 🚨

The Bigger Picture

And, as ever, this isn't going to land equally.

If your parents can support you through six months of applications, introduce you to someone at their firm, fund an unpaid internship or let you live at home while you build experience, a difficult graduate market is annoying.

If you don't have those things, it can become a wall. 🧱

That matters particularly for young people from lower-income and underserved backgrounds, because entry-level recruitment has historically been one of the few relatively structured routes into industries where personal networks otherwise matter enormously.

Then there's the wonderfully modern feedback loop: young people are applying for hundreds of jobs, increasingly using AI to help them apply, while employers increasingly use AI to screen those applications. 🤖🔄🤖

Everyone is automating. Nobody appears particularly happy.

So What?🧠

We need to stop treating entry-level employment purely as an efficiency problem.

A junior employee might be slower than an AI tool today. That's partly the point. They're learning. 🎓

Businesses that automate away their training pipeline may save money now and discover in five years that everyone wants experienced talent and nobody bothered creating any.

And for graduates, the message is unfortunately becoming clearer: a degree alone isn't enough differentiation anymore. Projects, internships, apprenticeships, portfolios, communities, practical AI skills and actual evidence that you can do something are becoming increasingly important.

The ladder isn't completely gone. But the first rung is getting considerably harder to reach. 🪜

📚 Read More

BBC News — Graduate job vacancies drop by almost 50% in a year, survey suggests
https://www.bbc.co.uk/news/articles/crerd8l3pw5o

💾Ramageddon

TL;DR: Amazon has hiked the price of some hardware by as much as 60%, blaming soaring memory and storage costs. Translation: the AI infrastructure boom is starting to show up in the price of ordinary gadgets  whether you asked for an AI revolution or not. 🤖💸

What Happened

Amazon has raised prices across Echo, Kindle, Fire TV and Eero devices, with the humble Echo Dot jumping from $49.99 to $79.99 overnight. Amazon says memory and storage component costs have risen so sharply that, after “absorbing these increases for as long as we could”, it is now passing them on to customers. 📈

And Amazon isn't alone. Apple has already raised Mac and iPad prices, Xbox has just made a six-year-old console considerably more expensive, and memory manufacturers expect shortages to remain painful through 2027 and potentially into 2028. Welcome to RAMmageddon 💾🔥

Why This Matters?

The culprit sitting behind much of this is AI. The world's biggest technology companies are buying enormous quantities of advanced memory to feed GPUs and data centres, and manufacturers are naturally prioritising the parts with the highest margins and strongest demand. Unfortunately, your Kindle also needs memory. 📚🤖

So here's the slightly awkward bit: you don't actually need to use AI to pay for the AI boom. The enormous infrastructure race between OpenAI, Google, Meta, Amazon and others is competing for components used throughout consumer electronics, pushing costs through the wider supply chain. Congratulations, you've been enrolled in the AI economy anyway. 🫠

The Bigger Picture

This is a useful reminder that AI isn't just software living mysteriously “in the cloud”. Behind every prompt are chips, memory, power, data centres, cooling systems and an increasingly ferocious fight for physical resources. ☁️➡️🏭

And there are distributional consequences. A 60% increase on a smart speaker is irritating if you've got money. Rising laptop, phone and console prices are much more consequential for students, families and people already struggling to afford the devices increasingly required for education and work. The digital divide gets rather harder to close when the hardware keeps getting more expensive. 📱⚖️

There is some hope: memory manufacturers are investing billions in additional capacity, while companies such as Google are developing techniques that allow AI systems to use memory more efficiently. But new semiconductor factories take considerably longer to build than another AI startup takes to announce a $10bn compute deal. 🏗️

🧠 So What?

We've spent years hearing about how AI will make products better, cheaper and more efficient. Right now, one of its more noticeable consumer achievements appears to be making an Echo Dot cost 60% more. Lovely. 😂

More seriously, this is the hidden AI tax worth watching. Data-centre demand doesn't stay inside data centres. It affects chips, electricity, land, water and increasingly the price of the devices sitting in your house. 💸

The AI race may eventually deliver enormous productivity gains. In the meantime, apparently Alexa would like £20 more please. 🔊

📚 Read More

TechCrunch — Amazon hikes hardware prices by 60%, blaming memory shortage
https://techcrunch.com/2026/08/24/amazon-hikes-hardware-prices-by-60-percent-blaming-memory-shortage/

TechCrunch — Memory chip giant SK hynix could help end ‘RAMmageddon’ with blockbuster US IPO
https://techcrunch.com/2026/03/27/memory-chip-giant-sk-hynix-could-help-end-rammageddon-with-blockbuster-us-ipo/

💸A big settlement💸

TL;DR: TikTok has agreed to pay $400m to settle US allegations that it illegally collected data from children under 13. It’s one of the largest child-privacy settlements ever and a reminder that when platforms say “our service isn’t for kids,” regulators increasingly want proof. 📱

What Happened

The US Department of Justice sued TikTok and ByteDance in 2024, alleging they collected “vast amounts of data” from millions of children under 13 without proper parental consent, in breach of COPPA — the Children’s Online Privacy Protection Act. TikTok will pay $300m immediately, with another $100m due when an earlier FTC consent decree is lifted. 💸

The case dates back to Musical.ly, TikTok’s predecessor, which had already paid $5.7m in 2019 for similar violations. Prosecutors argued TikTok was effectively “directed to children” while failing to properly verify ages or obtain parental consent. So this is less one unfortunate mistake and more we have apparently been discussing this for quite a while. 😬

And TikTok isn’t alone. YouTube previously paid $170m, Epic Games paid $275m, and Meta is currently facing a major child-privacy trial brought by 29 US states, with potential penalties that could run far higher. The regulator’s message is becoming pretty clear: children’s data is not just another growth metric. 🚨

Why This Matters

The interesting bit is that this cuts across the usual US-China framing around TikTok. Yes, TikTok has been at the centre of national-security concerns because of ByteDance’s Chinese ownership. But this case is much simpler: if you collect children’s data unlawfully, you can get hit whether you’re Chinese, American or anything else. ⚖️

That matters because child safety is becoming one of the few areas where regulators across the US, UK and EU are increasingly converging. Governments may disagree on AI, competition and free speech, but “maybe don’t quietly profile 12-year-olds” is proving remarkably bipartisan. 👶

The Bigger Picture

The deeper issue is that many social platforms were built around maximising engagement first and solving age verification later. Datamaxxing maybe?

That model gets awkward when the users being engaged are children and the product is also collecting behavioural data for recommendation systems and advertising. 📊

And this is where the business incentives matter. Younger users are valuable because they become long-term users. Platforms therefore have a powerful incentive to grow with them — while regulators increasingly expect those same companies to prove they aren’t exploiting them. That tension isn’t going away. 📈

🧠 So What?

A $400m fine sounds huge, but ByteDance was last valued at around $550bn. For companies at this scale, the real question is whether fines genuinely change behaviour or simply become another line item under “regulatory costs”. 💰

The bigger shift is cultural. Regulators are moving from asking platforms to have policies to asking whether those policies actually work. “Under-13s aren’t allowed here” means very little if millions of under-13s are still on the platform and their data is still being collected. 🧐

And for parents, schools and policymakers, this is the uncomfortable reality: the burden of protecting children online is finally being pushed back towards the companies designing the systems, rather than simply telling families to supervise harder. Probably about time. 🧠

📚 Read More

BBC News — TikTok to pay $400m to US in one of largest child privacy settlements
https://www.bbc.co.uk/news/articles/cwyr0l45xjro

👗Not so fast fashion👕

TL;DR: Shein is targeting a valuation of roughly $27bn for its Hong Kong IPO  a dramatic comedown from the $100bn investors once assigned it in 2022. Turns out ultra-fast growth looks slightly less magical once tariffs, regulation, slowing sales and actual scrutiny turn up. 👀

Yep Shein plans to list in Hong Kong on 1 September, offering nearly 280 million shares and hoping to raise up to $1.77bn. At the top of its pricing range, that would value the company at about $26.8bn. 💸

That sounds enormous until you remember Shein was valued at around $100bn in 2022. So yes, still worth billions. But also: congratulations on discovering what happens when a pandemic-era growth story meets reality. 📉

The company’s numbers help explain the haircut. Sales growth slowed to 8% in 2025, down from 20% the year before, and Shein reported a $99m quarterly loss earlier this year. The US ending its de minimis exemption, which had allowed low-value imports to avoid duties has also hit the economics of Shein’s ultra-cheap model. 🧾

Shein’s real superpower was never just “cheap clothes”. It was a supply chain and app experience designed to turn trends into products incredibly quickly, then serve users an almost endless feed of items cheap enough to buy on impulse. 📱

Which is why the FT makes a slightly more interesting comparison: maybe Shein isn’t really H&M with better logistics. Maybe it’s closer to a social platform with clothes attached. Scroll, dopamine hit, purchase, repeat. 👗📲

That framing matters because traditional fashion companies are judged on stores, inventory and margins. Shein behaves more like a digital attention machine, using data and gamification to keep people browsing and buying. Its competition isn't only Zara or Primark — it's arguably anything competing for your screen time and impulse spending. 🎰

The problem is that both sides of Shein’s story are now under pressure.

As a retailer, its price advantage is narrowing as the US and EU tighten rules on low-value imports, with the UK expected to follow. As a growth platform, revenue is slowing and profitability is weakening. Meanwhile, its attempted listings in New York and London both stalled, with regulatory and supply-chain concerns following the company around like an unwanted accessory. 🧳

And then there are the ethical questions that have never gone away: allegations around labour conditions, environmental impact and counterfeiting have all made investors more cautious. Shein says it has zero tolerance for forced labour, but public scrutiny has clearly become part of the valuation conversation. ⚖️

🧠 So What?

This feels like a useful reminder that private-market valuations are not commandments handed down from a mountain. They are guesses made during a particular moment, with a particular amount of hype, cheap capital and optimism. 🏔️

Shein at $100bn was a bet on endless growth, frictionless imports and continued dominance of ultra-fast fashion. Shein at $27bn is the market saying: maybe tariffs exist, maybe regulators matter, maybe growth slows and maybe consumers don’t scroll forever. Revolutionary stuff. 😂

The more interesting question is what kind of company Shein wants investors to believe it is now.

If it pitches itself purely as a fashion retailer, the valuation looks demanding compared with established rivals. If it pitches itself as a gamified commerce platform built around attention, data and habit, the comparison set suddenly starts looking more like gaming, social media and digital marketplaces.

Either way, the market is about to decide how much a dopamine-powered wardrobe is really worth. 👀

📚 Read More

BBC News — Shein aims for almost $27bn valuation in stock market debut
https://www.bbc.co.uk/news/articles/cdeweewjdxno

Gizmodo — Shein’s IPO Plans Value It at About $27 Billion. That’s Shockingly Low
https://gizmodo.com/sheins-ipo-plans-value-it-at-about-27-billion-thats-shockingly-low-2000802610

Financial Times — Shein’s IPO pitch should be more Meta than H&M
https://www.ft.com/content/0feb9a1f-a66a-45cf-a90d-bf0b43ab97bf?syn-25a6b1a6=1

📈 The tools behind the tech📉

📦Product📦

• How to think about token pricing
• AI usage patterns in software teams
  

📏Design📏 

• Why Tiny JPEGs Look Different in Chrome 

👩🏿‍💻Code👩🏿‍💻

• Software Engineering fundamentals matter more than ever

• Why Go is an Ideal Language for AI-Assisted Software Engineering

🏢The business behind the tech🏢

• How ikea comes up with its names

🙌🏾A Passport to Personalisation: Earning Customer Trust in AI🙌🏾

Ever wondered how AI knows you so well?  That's exactly what our next event is about. 
We are hosting a panel discussion all about AI personalisation and discussing how brands are navigating data, bias, and trust in the age of AI @ Tripadvisor

We have an incredible line up of speakers including: 
Shubham Kadlag: Manager, Marketing Engineering @ TripAdvisor
The person behind the tech that personalises how TripAdvisor reaches travellers across search, social, and CRM.
Subomi Salami: Senior Product Manager @ Microsoft AI
Senior PM at Microsoft AI, where she builds AI-powered experiences that enhance the way people work. 
Daniel Hammond: Senior Product Analyst @ Just Eat Takeaway
Co-founder of Just Eat's AI Ambassador Programme, championing knowledge sharing as the key to scaling with AI.
Tasneem Bhamji: Digital Engagement Director @ Lloyds Banking GroupAn executive at Lloyds Banking Group where she specialises in digital engagement and customer experience. 

There will also be giveaways and opportunities to network with people from across the tech community. 
📆Date: Wednesday 9th September
⌚️Time: 6pm - 9pm (with time for networking) UK Time
📍Where: Central London
Register to secure your free ticket here (spaces are limited and subject to availability) 

💴A chance to earn money💴

Tiiny host, the simplest way to share your work online is building a LinkedIn ambassador programme.

It's opened to UK/US creators with at least 1,000 followers, who are interested in AI workflow, who are builders at heart and love sharing their experience. If you love building landing pages, apps on Claude, Chaptgpt and would like to share your work, Tiiny Host is the perfect solution.

The opportunity is to promote Tiiny Host use cases and it's paid per post with 4 posts a month / 1 post per week.

More about the programme. You can apply here
Tiiny Host is offering a 20% discount off your first month using code promo colorintech to all Colorintech members

😃What we are consuming😃

🏡Remote workers report the highest well-being in study of 7,700 employees

🎤Mic Drop, a real-time multiplayer karaoke game