Over the last century humans have fallen in love with ideas, especially technology trends. We're addicted to the feature, and overestimate the short term impact while underestimating the long term impact.

Our team has been actively using AI daily for nearly 2 years, and it is a powerful tool. It's a much better search engine, great at repetitive tasks, speeds up productivity, and often very very wrong.

You've probably done a Google search and found the AI summaries useful, or used it to remove something from a photo, or maybe had it draft an email response. It's good, fine, ok, maybe bad at each of those things.

But one area I've gone heavy on is programming, following amazingly smart, funny and honest nerds like ThePrimeagen, and the other end of the spectrum some of the smartest most successful investors on the all-in podcast. I like to get my perspectives from a wide array of people.

Programming has been said to be one of the strongest areas in AI, and that's because it can be proven, tested, right or wrong. The app breaks, because something isn't right. That's very different to say writing, research, art, making a recipe, helping with paretning tips, because those are are not true or false, they are very complex and it's hard to know if it's right or wrong. Often it feels right just because it validates your perspective, opinion or in somecases is completely sycophantic.

My practical experience has been that it can help me learn a new programming language, Go, and how to work with Postgress, and then advanced my ability and speed in HTML, CSS and JS. However, when it's working in an area that I'm less familiar (Go) it works well, until it just doesn't, i.e. deleting databases, directly disobeying your rules, and requests, running illegal commands that share credentials publicly, completely undoing work it JUST DID when not requested.

The solution to these problems is a) you have to understand the subject matter b) give it very clear direction and c) monitor, intervene and review everything.

I say all this just to set the context, I believe it's a great tool, it's currently very flawed, but still valuable.

I am not buying the hype that it's capabilities will expoentially improve over the next 10 years, I am working with what I'm able to use and test myself today.

The hype is ridiculous, actually no it's completely expected

Humans love a good narrative, a story with a victim, villian and a saviour. We also can't help but be sucked in by fear. These things aren't news.

But what's happening now is a repeat of past hype cycles like this. Calculators. The loom. ATMs. Nuclear Power vs. Fossil Fuels. During each of these cycles there was insane hype that the world was about to change significantly for ever, and it's going to be bad.

Each time, things just kinda got better. ATMs were going to close branches, instead more bank branches opened. We gave people calculators and computers and hired more people to use them.

There are two common claims I'm seeing that are 'evidence' this time is different and the world order is changing.

1. Incumbent industries and the stock market is crashing because of AI

Every day there's another headline, social media post, podcaster claiming "See company X's stock price crashed because AI did a thing'.

I was getting sucked in and then had a moment of... wait let me check that myself. And in the major claims I've looked at they've proven to be false.

The Legal Sector (Feb 3: Anthropic Legal Plugin)

• Analysis: The charts show Thomson Reuters (TRI) dropping to the low $80s before rallying back to $103.11, and Relx (RELX) dipping but recovering to $34.76.
• The Reality: established players like LexisNexis and Thomson Reuters have massive, proprietary data moats and deeply entrenched enterprise contracts. An AI plugin is a tool, not an immediate replacement for a heavily regulated industry's infrastructure. LegalZoom’s (LZ) continued slide down to $6.29 suggests deeper business model vulnerabilities or broader market headwinds, rather than just a reaction to one Anthropic announcement.

The Security Sector (Feb 20: Claude Security Capabilities)

• Analysis: Crowdstrike (CRWD) took a sharp dive but has clawed its way back up to $442.03. Cloudflare (NET) not only recovered from its $177.14 dip but surged to $213.00.
• The Reality: Cybersecurity requires zero-trust architecture, compliance certifications, endpoint agents, and massive network infrastructure. AI makes these companies better, and they are integrating their own AI rapidly. A new capability from an LLM provider doesn't replace an enterprise firewall or endpoint detection system overnight.

Legacy Tech (Feb 23: Claude COBOL Announcement)

• Analysis: IBM was already in a steep multi-week slide from the $300s down to the $220s before the February 23rd announcement. It then actually rebounded to $248.87.
• The Reality: Modernizing COBOL is a massive pain point for banks and governments, but you can't just plug an AI into a mainframe and click "update." These are multi-year, highly complex migration projects. If anything, AI tools will likely be used by IBM consultants to speed up their own service delivery.

This is a well-documented phenomenon

that sits at the intersection of technology, psychology, and finance.

While there isn't one single name for it, investors and technologists usually refer to it through a combination of a few key concepts

1. Amara’s Law (The Hype Cycle)

Coined by Roy Amara, a researcher and scientist, this law states: "We tend to overestimate the effect of a technology in the short run and underestimate the effect in the long run." When a new technology drops, the media and market extrapolate its ultimate, futuristic potential and price it into the present day. This creates the massive, immediate spikes for the new tech and the corresponding panic-drops for the incumbents.

2. The "Disruption Myth" & Competitor Panic

In the stock market, this manifests as a knee-jerk "competitor panic." Algorithmic trading bots are programmed to scan news headlines. When they see "Anthropic releases legal AI," they immediately short or sell established legal tech stocks. Retail investors see the drop, panic, and follow suit.

However, they temporarily forget about Economic Moats and Switching Costs. Enterprise companies (like hospitals, banks, and law firms) don't buy software based on a shiny demo video; they buy based on compliance, security, existing integrations, and liability. Once the initial headline panic fades, institutional investors realize the incumbent isn't dying tomorrow, and the stock price corrects.

Historical Precedents

This exact cycle has played out repeatedly over the last century.

Cloud Computing vs. Legacy IT (2010s)

• The Hype: AWS and SaaS companies like Salesforce were going to immediately bankrupt "on-premise" legacy giants like Microsoft and Oracle.
• The Panic: Microsoft's stock was famously stagnant for years, often dipping when AWS announced huge growth, as people assumed legacy enterprise tech was doomed.
• The Reality: Enterprise systems are deeply entrenched. Microsoft used its massive cash flow and existing enterprise relationships to pivot, building Azure and shifting Office to the cloud, ultimately becoming more valuable than ever.

Electronic Spreadsheets vs. Accountants (1979 - 1980s)

• The Hype & Panic: When VisiCalc (the first electronic spreadsheet) and later Lotus 1-2-3 were released, there was widespread panic that bookkeepers, clerks, and accountants would be wiped out. A piece of software could suddenly recalculate thousands of cells in seconds—a process that previously took teams of humans days with paper ledgers and adding machines.
• The Reality: The spreadsheet didn't destroy the accounting profession; it expanded it. Because the cost of running complex financial models plummeted, the demand for financial analysis skyrocketed. Routine clerical jobs shifted, but the number of accountants and financial analysts actually grew massively. The software became a lever, not a replacement.

Nuclear Power vs. Fossil Fuels (1950s)

• The Hype & Panic: In 1954, the chairman of the US Atomic Energy Commission famously predicted that nuclear energy would soon be "too cheap to meter." The prevailing narrative was that oil and coal were dead industries walking, destined to be replaced overnight by clean, limitless atomic power.
• The Reality: The market radically underestimated the physical, regulatory, and financial realities of building nuclear infrastructure. Safety protocols, massive capital requirements, and eventually public backlash (like Three Mile Island) slowed adoption to a crawl. The fossil fuel industry didn't die; it went on to experience massive, multi-decade boom cycles.

The ATM vs. Bank Tellers (1970s - 1980s)

• The Hype & Panic: When Automated Teller Machines (ATMs) were rolled out en masse, it was considered the ultimate "job killer" for the banking sector. The media and unions assumed human bank tellers would be entirely automated out of existence within a decade.
• The Reality: The exact opposite happened. Because ATMs made it significantly cheaper to operate a bank branch, banks started opening more branches. While the number of tellers per branch dropped slightly, the sheer volume of new branches meant the total number of bank teller jobs actually increased between 1980 and 2010. Their roles just shifted from dispensing cash to selling credit cards and loans.

The VCR vs. Hollywood (1980s)

• The Hype & Panic: In 1982, the head of the Motion Picture Association of America (MPAA) lobbied Congress to ban the VCR, famously testifying: "I say to you that the VCR is to the American film producer and the American public as the Boston Strangler is to the woman home alone." The fear was that no one would ever go to a movie theater again if they could watch movies at home.
• The Reality: Home video didn't kill the film industry; it became its most lucrative secondary revenue stream in history. The VHS tape birthed empires like Blockbuster, created the "direct-to-video" market, and enriched the very studios that tried to destroy it.

2. We're all going to lose our jobs

The greatest theat to us all driving fear and politics "The jobs. We need more jobs" and AI is going to take that all away from us. Now sure there will be many jobs that will be changed, and repetitive, simple tasks are already being impacted, no dobut. And I am genuinely concerned for the least educated, empowered people who will likely be the most hurt by this, not people like myself. 

However, the hype currently is unfounded, it's on a dream or an idea, that a) AI will work and b) we wont make new jobs for people to do different things.

A recent 'evidence' this is already happening. Was Block firing 40% of their staff because of AI. Now this assumes that a) tech companies are completely honest, and b) they don't need to improve shareholder value while use a trend to cover their mistakes....

Over the last many months and years, specific to AI we've seen complete 180's - "We're cutting our staff because of AI" then months later "Oh crap we need you all back sorry". To more hilarious examples where Anthropic (and others) keep saying programming will be 100% in 18 mths, 12 mths, 6 mths, 3mths, wait, not quite yet.. all the while they continue to hire record numbers of developers.

Again, this whole trend isn't new.

Here are four examples where tech companies used a "major technological shift" as cover to execute massive layoffs, only to rehire heavily into the new paradigm shortly after:

1. Microsoft: The "Cloud First" Shift (2014)

• The Cut (July 2014): Microsoft announced 18,000 layoffs (about 14% of its workforce). While much of this was the fallout of the failed Nokia acquisition, CEO Satya Nadella explicitly used the cuts to frame a broader restructuring around a permanent, structural shift away from legacy desktop software toward a "Mobile First, Cloud First" world.
• The Rehire (2015-2018): Microsoft used the cover of this shift to clear out legacy management and low-growth divisions. Almost immediately, they went on an aggressive hiring tear for Azure cloud engineers, data scientists, and enterprise sales. By 2018, total headcount had grown past its pre-layoff peak (from ~128,000 in 2014 to ~131,000 in 2018).

2. Intel: The "PC to Cloud" Transition (2016)

• The Cut (April 2016): Intel cut 12,000 jobs (11% of its workforce). The stated reason was a permanent market shift: moving away from the "dying" PC market to focus on the booming cloud computing and connected devices (IoT) sectors.
• The Rehire (2017-2019): The layoffs boosted margins and appeased investors worried about PC sales. Meanwhile, Intel quietly and aggressively recruited data center engineers and AI hardware designers. By 2019, its global headcount had surged back to over 110,000, significantly higher than the 106,000 employees it had immediately following the 2016 cuts.

3. Meta (Facebook): "The Year of Efficiency" vs. AI (2022-2023)

• The Cut (Nov 2022 - Mar 2023): After massive overhiring during the zero-interest-rate era, Meta cut 11,000 jobs, followed by another 10,000. Mark Zuckerberg framed this as the "Year of Efficiency," blaming a permanent macroeconomic shift in digital ads and e-commerce.
• The Rehire (Late 2023-2024): While the cuts instantly improved the stock price, Zuckerberg quickly pivoted the narrative to state that Meta's single largest investment moving forward would be AI. By late 2023, Meta was aggressively opening new requisitions and offering massive compensation packages for AI researchers and infrastructure engineers, effectively backfilling engineering capacity under a new technological banner.

4. IBM: The "End of the Mainframe" (1993)

• The Cut (1993): Under CEO Lou Gerstner, IBM initiated unprecedented rolling layoffs that eventually eliminated roughly 60,000 jobs. The market narrative was that the era of mainframe computing was structurally dead, killed by the PC revolution.
• The Rehire (1995-1999): Gerstner used the "dying hardware" narrative to completely clean house and pivot the company to IT services and software (IBM Global Services). Within a few years, IBM was hiring tens of thousands of consultants. The company didn't shrink long-term; it used a technology shift to swap out its workforce and reset its baseline without admitting the previous decade of management was flawed.

The end isn't nigh

My opinion is it's going to change the world, like previous technologies have, but let's put the hype aside and deal with reality, find a positive path forward for everyone. And don't be scared, it's going to be ok, and we'll figure it out, like we have for centuries.