Once again, it’s time for Intellyx’s annual predictions, a tradition dating back to the beginning of the millennium. We don’t simply throw our predictions out into the void, however; we actually take turns with our annual prognostications and review the successes – and failures – of the analyst who stuck his neck out the year before.
This year I take my colleague Jason ‘JE’ English to task. Were his three predictions from this January 3, 2025 Cortex on target, or did he miss the mark?
JE’s predictions from a year ago
JE first predicted that bespoke architecture would return. With cloud native architecture commoditized, he thought that enterprises would seek out more opinionated stacks based on targeted business use cases.
I’m giving JE a grade of C on this one, because he failed to place the trend to bespoke architectures into the realm of AI.
It has indeed come to pass that enterprises are building opinionated AI stacks to target specific business problems, as well as to avoid public LLMs.
Specialized language models (SLMs) are the trend du jour – more efficient, less likely to hallucinate, and easier to run on premises or in a private cloud instance.
Outside of architectures for supporting AI, however, I haven’t noted a significant trend to bespoke architectures in 2025. Even edge computing, which lends itself to bespoke architectures, has focused primarily on AI-supported use cases.
JE’s second prediction was that the dearth of solid AI adoption and governance policies going on at both vendors and enterprise end customers would change drastically, given the high-profile failures of rogue AI usage.
Sorry, JE, I’m giving you a C on this one as well. While AI governance is an increasingly important priority for enterprises, the challenges facing AI adoption have only grown.
Shadow AI in particular is exploding, and while several vendors are tackling the problem, it is still on the upswing.
JE’s third prediction was also premature: that quantum computing would break cryptocurrency. The fact that quantum computing threatens classical cryptography is undeniable, as I wrote about last May – but the timeframe for quantum supremacy is still a few years off.
The cryptocurrency world, furthermore, is apparently on top of the problem as well, although the rollout of quantum-safe private keys won’t protect anyone who doesn’t upgrade their wallets.
In other words, the scammers are hoping to protect their own crypto while picking off the wallets of unsuspecting marks – but not for a few years yet.
My predictions for 2026
While I’m on the subject of cryptocurrency, I can’t help but make a crypto prediction. Will Bitcoin crash this year? True, all crypto is a Ponzi-like scam, as I explained in a 2021 Cortex, but predicting that the inevitable crash will occur in 2026 would be too easy. I’d rather make a guess that other people aren’t making.
Bitcoin ATM hack
My prediction: there will be a massive organized crime takedown of Bitcoin ATMs using counterfeit currency and SMS spoofing.
Bitcoin ATMs’ primary purpose, of course, is fleecing gullible seniors of their retirement savings. My prediction turns the tables, making the ATM companies themselves the victims.
Even though ATMs validate deposited currency using multiple tools, skilled counterfeiters are able to defeat these measures – especially if they reverse engineer the devices.
The reason the traditional ATM network is resistant to such attacks is because banks are able to check deposited currency manually, and in any case, deposits can be reversed if necessary.
But not in the case of Bitcoin ATMs. Once the machine writes the deposit to the blockchain, it’s both irreversible and also difficult to track. The money might be anywhere in the world.
Furthermore, most Bitcoin ATMs will accept deposits of up to $2,000 with only an SMS confirmation – no further ‘know your customer’ (KYC) required.
Major agentic AI financial attack
We’ve already seen one instance of hackers using agentic AI to mount a cyberattack. AI vendor Anthropic was able to stymie alleged Chinese state hackers who used Claude to infiltrate several large organizations.
My prediction takes this trend one step further: I predict a similar attack will compromise one or more financial institutions to steal millions of dollars – or more.
This ‘theft of the century’ will strike fear in the hearts of bankers around the world as they realize they are increasingly vulnerable to innovative agentic attacks.
Critical thinking crisis
The more people use generative AI (genAI) to do their jobs, the less able they are to do those jobs themselves, as I explained in my last Cortex.
In other words, genAI eventually turns developers and other knowledge workers into useless boneheads.
This problem is even worse in educational institutions around the world. Students no longer have to think for themselves. AI can do their homework for them. The result? More useless boneheads.
You’d think, then, that HR departments would be able to separate such boneheads from people who are actually qualified for the jobs they seek.
Sorry, no. It’s now so dead simple to use genAI to craft resumes, cover letters, and even interview training material that yes, any useless bonehead can do it.
What are recruiters to do? Use AI to review resumes, conduct interviews, etc. In other words, boneheads interacting with boneheads.
My prediction: this rise of boneheadedness will lead to a critical thinking crisis in 2026.
The easier AI is to use, the worse it will get, until executive management finally sounds the alarm – unless it’s already too late, and the C-suite is already filled with, you guessed it, useless boneheads.
The Intellyx take
OK, I admit it, I’m a grumpy old curmudgeon. All three of my predictions are decidedly negative. Wasn’t I able to come up with even one prediction with a positive outlook?
I’m sure I could have if I tried harder – but what’s the point? For this Cortex to be useful, it should highlight problems that might fall into people’s blind spots – and the bigger the blind spot, the bigger the potential problem.
You might say that I’m fighting a Sisyphean battle against boneheadedness.
After all, I’m from the generation who wrote all our term papers on a typewriter, one letter at a time. Critical thinking was job number one – a hard-won skill from our formative years we took into our careers.
Today, AI is doing the thinking for us – for better or worse, mostly worse. My advice for reversing this trend? Don’t be a bonehead.
Copyright © Intellyx BV. Intellyx is an industry analysis and advisory firm focused on enterprise digital transformation. Covering every angle of enterprise IT from mainframes to artificial intelligence, our broad focus across technologies allows business executives and IT professionals to connect the dots among disruptive trends. No AI was used to write this article. Image credit: Craiyon.