Every week there is another headline warning that the AI bubble is about to burst and cause a total market collapse.

To be honest, “crash” is just a lazy doomsday word people throw around for clicks.

What we are actually looking at is an inevitable market correction. There is a massive difference between stock prices snapping back to real-world value and a technology disappearing overnight. AI is here to stay, but the economics behind the current hype cycle simply do not add up.

Here is my take on what is really happening beneath the surface, why the maths is broken, and what happens when the bill finally comes due.

The Dot-Com Boom Parallel: We Have Seen This Movie Before

To understand where tech is heading today, you have to look back at the dot-com boom of the late nineties.

Back then, the long-term vision for the internet was completely right, but the timing, costs, and infrastructure were completely wrong. Everyone rushed to build websites, yet hardly anyone had the home broadband or mobile access to actually use them properly. You had online pet food shops and e-commerce startups trying to sell goods before nationwide parcel delivery and modern digital payment rails even existed.

When the market correction hit, hundreds of overvalued dot-com startups vanished into thin air. But did the internet die? Not at all. Amazon, eBay, and the foundational pipes of the web survived, matured, and went on to run the modern world.

That is exactly the cycle the AI sector is walking into right now.

The Data Centre Black Hole and Tech Circular Financing

The amount of capital expenditure being poured into AI infrastructure right now is mind-boggling. We are talking hundreds of billions, pushing towards a trillion dollars globally in chips, massive server farms, and energy grids.

Yet if you look at the balance sheets of most AI-first startups, the actual returns are tiny compared to what is being spent.

Think of it like building a luxury block of flats that cost millions to put up. You manage to rent out a couple of spare rooms for pocket change, and you are nowhere near breaking even on the mortgage, let alone turning a profit.

Right now, a handful of mega-cap tech giants are essentially propping up the entire market. In many ways, it is the same pot of cash circulating in a closed loop: Big Tech invests in AI startups, and those startups immediately hand that cash right back to Big Tech to pay for cloud computing and GPU clusters.

When private equity, venture funds, and banks hand over borrowed billions, they expect a tangible return within three to five years. Eventually, investors are going to stop nodding along to hype presentations and ask for their money back with actual interest. When they demand companies put their money where their mouth is, the music stops.

The Subsidised Reality: We Aren’t Paying the True Cost of AI

Ask yourself a simple question: how much does running these massive models actually cost?

Right now, everyday users pay around £20 a month for ChatGPT Plus, Claude, or Google One AI. But that price is heavily subsidised by venture capital and corporate cash reserves burning through runway.

Between the raw electricity, cooling, water usage, and GPU wear-and-tear, running thousands of complex reasoning queries a month costs significantly more than £20 per head. For these providers to break even and turn an honest profit, that subscription could easily need to be closer to £80 or £100 a month.

It reminds me a lot of the lead-up to the 2008 subprime mortgage crisis. Cheap money was handed out freely to inflate the market and get everyone through the door, hiding the structural rot underneath.

If tech giants are forced to price these tools at their actual operating cost, how many casual users and small businesses will keep paying? The market will shrink rapidly to only those who get genuine, measurable enterprise value from it.

Open-Source AI and Global Efficiency

To make matters worse for the big spenders, the idea that you need infinite capital to build useful AI is already falling apart.

Teams in China and the global open-source community are releasing compact, highly efficient models that achieve 80% to 90% of the capability at a tiny fraction of the training and inference cost.

If lean, open-source models can run locally or on modest servers, why would the market indefinitely fund bloated, monstrously expensive proprietary data centres? Building endlessly expensive hardware setups for diminishing returns makes no commercial sense long term.

Market Outlook: Who Actually Survives the Correction?

When the market correction lands, it is going to hurt. Stock prices will take a heavy tumble, and dozens of wrapper startups that never had a sustainable business model will shut their doors.

In the end, only two groups will walk away intact:

1. The mega-caps with bulletproof balance sheets: Companies that generate reliable billions from other core businesses (like enterprise software, search ads, and retail) and can afford to absorb the write-downs.

2. Critical sovereign infrastructure: Models and platforms backed or subsidised by governments for defence, cybersecurity, and national data sovereignty.

AI is not going to vanish, but the era of easy money, cheap subscriptions, and unchecked data centre spending is running on borrowed time.

Don’t know about you, some believe a mega crash, end of AI. Other’s believe it is all fear mongering. I’m in the middle.

It has been a good while since my last post on here.

Between sorting out bits on the house, keeping on top of cybersecurity at work, and just general day-to-day life, I have not had the time to sit down and write. Things have calmed down a bit now though. I finally have the headspace to get back to what I enjoy doing: working on projects, testing gear, and sharing my thoughts on where tech is heading.

Here is something that caught my eye today.

The Tamagotchi Throwback: What Is the Meta Muse Charm?

Meta has just shown off a new pocket gadget called the Muse Charm. It is tiny, has a little two-inch screen, comes with built-in 5G, and runs their autonomous assistant, Muse.

The first thing that crossed my mind when looking at it was how much it looks like a modern Tamagotchi.

Growing up as a millennial, everyone had one. It was that little plastic egg you clipped to your bag or keys, checked in on, fed, and tried to look after before you forgot about it and it died.

Gen Z never had that firsthand. Yet right now, you see the younger generation leaning hard into late nineties and early noughties culture. Baggy clothing, wired earphones, and older iPods are making a massive comeback. A lot of people are actively trying to cut down their screen time and step back from social media feeds.

From that perspective, packaging an AI assistant into a retro-style physical toy is quite a clever angle from Meta.

The Smartphone Screen Time Trap

Think about what happens whenever you want to check something quickly.

Say you want a quick recipe idea for dinner. You take your phone out, unlock the screen, and before you even open a search tab, you see unread WhatsApp notifications, calendar alerts, and banners popping up.

Next thing you know, you are replying to messages and checking updates. You completely forget what you picked the phone up for in the first place, when all you wanted was a simple answer to what to cook.

The appeal of a single-purpose companion makes sense here. You ask your question, get the answer, and carry on without getting sucked into your phone.

Dedicated AI Hardware: The Practical Hurdles

Making AI feel like an interactive pocket pet is definitely a smart way to appeal to younger users. It also links straight into WhatsApp, which is useful because even if people cut down on Instagram, WhatsApp is still how most of us keep in touch with family and friends.

That said, looking at it practically, there are a few obvious snags.

• Another device to carry: It has a little cord, but where do you actually put it? Clipped to a backpack? Hanging off your car keys? It is still another piece of hardware to remember and another battery you have to charge at the end of the day.

• The smartwatch issue: If the point is a small, glanceable screen that gives you quick answers without touching your phone, most of us already wear a smartwatch. Is digging a small charm out of your pocket really better than just tapping your wrist?

The Bigger Picture: Agentic AI and E-Commerce

There is a bigger picture here. Meta is not just making a cute novelty device for the sake of it.

Under the surface, Muse is built to take actions for you, like booking things or buying products online. Meta has already linked up with Shopify so the assistant can browse and purchase items on your behalf. Amazon has already moved to block Muse on their site because they know Meta is trying to position itself right in the middle between you and the retailer.

Whether people actually bother carrying a separate keychain gadget remains to be seen. But testing out hardware that gets people talking to AI without looking at a phone screen is an interesting move.

I’m wondering who would carry a dedicated pocket AI device to cut down on screen time, or is your phone and watch setup already doing the job?

You’re not alone and it might not be just high demand, inflation, or fuel costs to blame.

A hidden digital culprit is quietly inflating ticket prices: AI-powered bots.

In recent months, airlines have reported a significant rise in bot-driven attacks and it’s changing how flight prices are calculated and displayed online. These bots aren’t new, but recent advances in AI have made them faster, more scalable, and more disruptive than ever before.

🤖 How Are AI Bots Inflating Flight Prices?

According to security experts and bot monitoring reports, here’s how it works:

• Fake Bookings (Seat Spinning): Bots reserve flight seats in large volumes without purchasing them, causing airline systems to believe demand is high. This triggers price hikes.

• Scalping and Hoarding: Bots may even buy tickets en masse to resell at inflated prices via third-party platforms.

• System Slowdowns: Bots bombard airline websites, APIs, and notification systems (like SMS alerts), causing slowdowns and disrupting services for genuine customers.

The goal? Manipulate supply and demand so you end up paying more.

⚠️ Why It’s Worse Now

This kind of bot activity has existed for years, but it’s now supercharged by AI.

Thanks to generative AI tools and low-code bot platforms, even low-skilled attackers can launch complex bot campaigns. They use VPNs and fake browser profiles to avoid detection, making traditional defences like CAPTCHA increasingly ineffective.

The result: airline ticketing systems are overwhelmed, pricing algorithms get confused, and we all feel it when booking our next holiday.

💸 Who Is Most Affected?

• Families booking last-minute flights

• Budget travellers using comparison sites

• People browsing on mobile or at peak times

Even the device or browser you use can affect the prices you’re shown a tactic some airlines and travel platforms have been known to use.

🧠 What Can You Do?

• Use incognito/private browsing mode when searching for flights.

• Compare prices across multiple websites.

• Clear your cookies regularly to avoid personalised pricing.

• Track prices early and watch for fluctuations.

• Be cautious of third-party booking links - they might be part of a scam.

🌍 Final Thoughts

As someone who works in cyber security, I find it both fascinating and concerning how something as routine as booking a holiday can now be influenced by invisible AI-driven systems.

It’s a reminder that digital threats are no longer just about data breaches - they’re affecting our wallets too.

Let’s stay informed, help each other out, and share this knowledge. Your awareness could save you and someone else hundreds of pounds this summer.

🔗 Found this helpful? Subscribe for more tips on digital safety, tech trends, and cybersecurity advice that affects your everyday life.

🗣️ Feel free to share this with friends

That moment of convenience could now cost you.

QR code scams, particularly on parking meters/billboards are becoming increasingly common across the UK.

Fraudsters are taking advantage of our reliance on smartphones to create convincing fake stickers, tricking people into sending money directly to them.

🎥 The Scam in Action

In a recent video I came across, a driver was about to pay for parking when they noticed something odd. The QR code on the sign looked like a sticker. Upon closer inspection, they realised it had been stuck on top of the PayByPhone logo.

If they had scanned and paid, their money would have gone straight to a fraudulent website, not the local council. No confirmation. No receipt. Just gone.

🧠 How the Scam Works

• Scammers print fake QR codes that look identical to the original

• They stick them over real QR codes on parking machines

• Victims scan the fake code and are taken to a fraudulent payment page

• The payment goes to a scammer’s account, often via crypto or third-party wallet
  Some people receive a fake receipt, others get nothing at all

🚨 Why This Matters

This kind of fraud is hard to detect at a glance, especially when you’re in a rush to park.

These QR scams are showing up in car parks, high streets, and even hospitals. Most people don’t realise they’ve been tricked until they check their bank account later.

🔒 How to Stay Safe

• Avoid scanning QR codes directly on parking machines

• Instead, download and use the official parking app (e.g. RingGo, PayByPhone)

• Inspect QR codes carefully - If it looks like a sticker or is misaligned, be cautious

• Report suspicious QR codes to your local council

• Share this post with others - The more awareness, the fewer victims

📣 Final Thoughts

Scams like this work because they blend into everyday life. The best defence is being aware and taking a few extra seconds to verify what you’re scanning.

If you’ve seen similar scams or had a close call, leave a comment below - it could help someone else avoid the same trap!

📩 Subscribe to my blog for regular cybersecurity tips, tech updates, and real-world scam alerts

The caller simply asked, “Is this David Thompson?” Thinking it might be work-related, he replied, “Yes.”

The caller followed up with, “Are you from [CompanyName]?” Again, David confirmed.

Then, silence. The caller hung up immediately.

At first glance, it may seem harmless. But this was actually a successful vishing attempt — a form of voice phishing designed to confirm personal details and link them together for malicious purposes.

What Actually Happened?

This wasn’t just a cold call. It was a deliberate attempt to validate three key pieces of information:

• David’s full name

• His personal phone number

• His place of work

With those confirmations, the attacker now has a verified profile, one that can be combined with public data (like LinkedIn or company websites) and used in more sophisticated attacks.

This kind of verified information is extremely valuable and can be:

• Sold to data brokers for telemarketing or robocalls

• Leaked or traded on the dark web

• Used in targeted phishing or impersonation attacks

• Compiled into larger datasets for future social engineering campaigns

How to Avoid Falling for Vishing Attacks

Even people in tech or cybersecurity can get caught off guard. The call seems innocent, but that’s exactly why it works. Here’s how to defend yourself:

1. Don’t Just Answer, Ask Back

If someone calls and asks, “Is this [Your Name]?”, follow it up immediately with:

“Who’s calling, and what is this regarding?”

Flip the dynamic. If they’re asking about you, they should identify themselves first. Be curious, not courteous.

2. Use Call Screening Tools

Modern smartphones offer smart ways to deal with unknown callers:

• Google Pixel phones have Call Screen, where Google Assistant answers unknown calls, asks why they’re calling, and transcribes their response in real time.

• Apple’s upcoming iOS 26 (currently in Developer Beta 4) introduces a new Call Screening feature where Siri can:

• Automatically answer calls from unknown numbers

• Ask the caller to state their name and reason for calling

• Display a live transcription so you can decide whether to answer

This puts a layer between you and unknown callers. Most scammers and cold callers won’t go through the effort.

3. Try the Voicemail Filter Trick

Here’s a personal tip I use that works surprisingly well:

My voicemail starts with: “Hello?”

I pause, most bots or scammers start talking right away. Then I say:

“If this call is important, please leave a message. If not, I won’t return the call.”

It filters out a surprising amount of junk.

Final Thoughts

This wasn’t a phishing email or a malware attack. It was a simple phone call, and yet it successfully confirmed three valuable pieces of information.

These details can now be weaponised in future scams, social engineering attempts, or sold for profit.

Stay sharp. Just because a call seems routine doesn’t mean it’s harmless.

Question everything, especially unexpected calls.

💡 Have you ever received a call like this? Drop a comment or share this post to raise awareness.

🔒 If you found this helpful, subscribe for more real-world cybersecurity insights and tips.

It’s been quite some time since my last post, and for that, I apologise. Life has been rather hectic, primarily because I’ve been diligently preparing for the CISSP exam. Recently, I printed a comprehensive 600-page study guide from the Jason Dion Academy on Udemy. This has been my main resource, although staying on top of it has been challenging with life’s distractions and my recent trip to Amsterdam.

My CISSP Study Strategy: The Pomodoro Technique Approach

In the coming weeks, I aim to share my journey with CISSP preparation. I’ve adopted the Pomodoro technique, dedicating at least an hour each day to study. By maintaining this schedule, including weekends, I intend to achieve between five to seven hours of study weekly. This approach indicates it will take me approximately six months to thoroughly grasp the materials.

What to Expect: CISSP Exam Difficulty and Time Requirements

The CISSP certification exam itself is demanding, spanning three to four hours and requiring a thorough understanding of the complex cybersecurity concepts covered. As one of the most challenging information security certifications, proper preparation is essential for success.

Coming Up: CISSP Study Tips and Insights

Throughout this journey, I plan to share:

• Valuable CISSP study tips

• Key notes and concepts

• Insights from my learning experience

• Time management strategies for busy professionals

• Reviews of study resources and materials

I appreciate your patience and look forward to reconnecting and keeping you updated with my progress. Stay tuned for more insights into my CISSP certification adventure.

Last week, I started my CISSP journey, and honestly? I’m loving it so far.

But as I’ve dived into the regulations domain, I quickly realised something, there are way more regulations than I expected! It’s a bit overwhelming, and I know I’m not the only one who’s felt this way. 😅

Share Andy’s Tech Talks

Why CISSP Regulations Are Challenging

I didn’t know there would be 8 known regulations to understand. That completely caught me off guard. I thought there would just be a handful, but instead, I found myself staring at a list of names and acronyms that seemed endless.

• PCI DSS? Sure, that one makes sense.

• GDPR? We’ve all covered that in countless HR training sessions.

• But SOX, GINA, GLBA? Never heard of them!

• The unexpected mix of familiar and unfamiliar regulations really threw me.

It’s not just about memorising the names, it’s about understanding the context of each one. For example, knowing how regulations like PCI DSS tie into cybersecurity can be challenging, especially when they’re not part of your daily work.

And let’s not forget the acronyms, I’ve heard of FTCA and HIPAA, but the rest? They’re completely new to me! 😅

My Approach to Memorising CISSP Regulations

To keep my study sessions effective, I’m planning to use flashcards. They’re practical for drilling acronyms and key points. Plus, I’m thinking about creating visual aids to make the information more digestible, I might even get some help from Professor ChatGPT for that!

Right now, I’m working through Jason Dion’s Udemy course: ISC2 CISSP Full Course & Practice Exam. Honestly, it’s been a lifesaver. The structured lessons and practice exams keep me on track, and the course’s tone is engaging and easy to follow.

Learning from the Cybersecurity Community

If you’ve tackled the CISSP regulations, how did you handle memorising them? I’d love to hear your strategies!

Also, if you have any resources or study techniques that worked for you, drop a comment. I’m always on the lookout for new methods to make this journey a bit smoother.

Consistency Is Key in CISSP Preparation

One thing I’ve learned so far is that consistency is key. I dedicate an hour a day to studying, which helps me soak up the new information without feeling overwhelmed. My approach is simple: enjoy the process and stay curious. Progress might be slow, but it’s all about taking consistent steps forward.

Final Thoughts on CISSP Study Tips

Sometimes, the volume of information feels like a lot to handle, but I remind myself that perseverance will get me there. I’m determined to stay focused, keep pushing through, and find ways to make studying enjoyable.

If you’re on your own CISSP journey or working through another cybersecurity certification, I’d love to know how you’re tackling the challenges. Let’s share tips and motivate each other!

Subscribe for More CISSP and Cybersecurity Tips

If you’re interested in following my journey through the CISSP certification process and want more tips on cybersecurity training, hit that subscribe button!

💡 Keywords: CISSP, Cybersecurity, Certification, Study Tips, Flashcards, Exam Prep, Cybersecurity Training, Information Security, Cybersecurity Learning

✍️ Let’s Connect: Drop a comment with your thoughts or advice!

🚀 Follow for More: Stay updated on my CISSP journey and learn how to navigate the world of cybersecurity certifications!

Leave a comment

But things got real when my colleague sent me this article from The Independent (link). It confirms that Scattered Spider—the same hacker group that hit MGM Resorts—was behind the attacks, and once again, they gained access by social engineering the IT help desk.

As someone who worked on the help desk for 8 years, this hit hard. We used to joke about users falling for phishing emails—but now it’s the tech teams themselves being duped.

And sadly, I can see how it happens:

• Many MSPs and internal help desks have high staff turnover.

• The focus is often on speed and efficiency—not strict identity verification.

• Impersonating a colleague over the phone still isn’t seen as a serious threat by some.

• Outsourced support teams, often in different countries, may not always understand local business culture or the risk of blindly trusting someone who sounds authoritative.

It’s a wake-up call.

I’m now thinking I may need to produce some cybersecurity awareness training tailored specifically for IT help desks—with real-world examples like this. Not just for our own team, but maybe even for our MSP.

In fact, I’m considering reaching out to our MSP directly to ask:

• Do their help desk staff receive monthly cybersecurity training?

• Can they share completion certificates to confirm this?

Because if we don’t challenge and verify this now, we might be the next ones in the headlines.

This is something I’ll be bringing up in our next all-staff meeting.

Has your organisation done enough to protect its help desk?

It’s been a little while since I last posted, and I wanted to share a quick update on where I’ve been — and why I’ve decided to move Andy’s Tech Talks over from Beehiiv to Substack.

Subscribe now

🧠 Burnout & Hitting Reset

To be completely honest — I burnt out.

Between juggling full-time work, content creation, personal projects, and ongoing studying (I’m currently working towards my CISSP), things started to pile up. I found myself spending more time tweaking layouts and stressing about growth than actually enjoying the process of sharing what I’m learning.

Taking a step back helped me realise I needed to simplify and create in a way that feels more sustainable.

🔁 Why Substack?

While Beehiiv is a solid platform, I’ve found that Substack just suits me better at this stage. Here’s why:

• It’s easier to manage — the writing process feels straightforward and intuitive

• Discoverability is built-in — it’s easier to connect with like-minded people here

• It has more of a social feel — I can follow other creators in the cybersecurity space and interact more naturally

• More freedom to post casually — Notes make it easy to share quick ideas without overthinking

Substack feels less like a job, and more like a community space where I can share my journey, learn out loud, and connect with others along the way.

📚 What’s Coming Up

Now that I’m back in the flow, I’ll be sharing more content on:

• Breaking into IT and cybersecurity

• What I’m studying (including the CISSP and other certifications)

• Cybersecurity and tech news (explained simply)

• Lessons from the books I’m reading

• Honest reflections from my career and learning journey

🙏 Thanks for Being Here

If you’ve followed me from Beehiiv — I really appreciate you.

If you’re new — welcome to the community.

This is a space for anyone looking to get into tech or grow in cybersecurity — especially if, like me, you’re figuring it out one step at a time.

More to come — let’s grow together.

– Andy

The documentary will be more of a reporter type explaining the facts and history. Hopefully, no one will think I am biased! We’ll get over that hurdle when it gets published, I can imagine the comments 🤦‍♂️

Not sure when it’ll be ready, just three parts left:

• Thumbnails

• Voiceover

• Video Editing

Might be ready and polished sometime in February, in life doesn’t get in the way ✌️

-Andy

-Andy

Strangely, there’s no redirection option in Hostinger to just direct all traffic from the main site to Beehiiv, that’ll be easier isn’t 🤔

I’ve updated the DNS records again, let’s see how it goes! If it doesn’t work a second time round, I’m just going to contact beehiiv support 👍

Another test post for the site ✌️

-Andy

On the other hand, I still need to subscribe to a newsletter platform. I have researched to use ConvertKit or beehiiv and beehiiv looks better. Their free plan is much more budget-friendly, and it also comes with a free website builder!

Let’s see, as I’m not too sure if the website template will come out any good… This is a test article post to be used and rearranged around the site 😃

Wish me good luck!

Andy

Subscribe now