Time to Get Organized

An optimistic prepper's guide for the next two years of AI expansion.

It’s time to start thinking about your future because right now a lot of powerful people are thinking about it for you.

Nine months ago, a cycle began in corporate board rooms and C-suites: last year’s profits were reported, predictions of future revenue were made, and plans were laid to ensure year-end balance sheets show as much profit as possible, even with less growth.

Even in stable times no one can predict the future of the economy, it’s always a guess. But in this moment corporate leaders don’t have to guess about an immediate future of volatile trade policies and political change creating near constant uncertainty — about everything.

The mandatory corporate response to uncertainty is to lower costs and protect future profits. And that always includes reducing the workforce.

Here’s the timeline:

In Q1, departments or product lines are selected for reduction or elimination. High level managers are given cost targets to hit and asked to make recommendations on staffing. This is all done in strict confidence, in meetings of small, high-ranking groups.

The first phase is already done. It happens in March/April when a round of layoffs are announced. The second phase, usually a larger round, occurs in July, at the beginning of Q3.

The logic is, 1) don’t announce everything at once — it takes time to adjust to the new distribution of tasks and, 2) make sure the adjusted staff is functionally in place by the beginning of Q3 to insure balance sheet goals are met by the end of the year.

This is the normal historic cycle of a downturn. People are let go, more in some industries than others, a certain percentage retire and the rest are eventually absorbed into other forms of employment.

I am writing this in June 2025 and I know with certainty the scale of this will be very different in the age of AI.

Chart source: Peter Diamandis, Metatrends

1999

1999 concluded the first growth era of the Internet. After ten straight years of massive capital investment connecting businesses and homes, the demand for profits finally overcame the promises of the pitch deck and the soaring stock market, on fire with a new wave of online investors, crashed hard.

Over the next three years many companies closed, not just in tech. The news media locked in on the stock market like it was the nation’s heartbeat. When investment returned it came rapidly, as companies began reporting profits from huge productivity gains driven by connected computing that sped up record keeping, financial reporting, manufacturing, shipping, communications, and more.

But something strange happened: for the first time, employment didn’t come back at the same rate. When capital returns it normally creates jobs, but this time it didn’t. The baffled news media called it, with zero irony, “The Jobless Recovery”.

Productivity Gains

After the dotcom bust when companies finally started hiring again salaries were lower because many specialized skills requiring expensive training had been automated by new software from Oracle, Intuit, Microsoft, Google, and Adobe. The jobs weren’t coming back because the US economy didn’t need them anymore.

Of course, plenty of manufacturing jobs were created. But thanks to new trade agreements and international coordination made possible by the Internet, those jobs were overseas where labor costs were comparatively minuscule. 25 years later, most of those jobs are still gone.

Eventually, the US did create higher paying jobs in a different kind of manufacturing — building applications for the Internet. Fifteen years after the crash, people with digital skills were making more than $100k a year as middle-managers for new businesses in logistics delivery, e-commerce, social media, and ad-tech. Hard skills paid a lot more: a starting salary for a mid-level coder could be around $200k. Advanced programmers who could build and manage processes for enterprise applications made millions.

In 2015 I started building applications using the first computational engines available via API. Back then they called it “cognitive” computing.

A note about my career: I’m a college dropout and had crap grades throughout high school. I won a photography scholarship that paid for a couple semesters of college but when the money ran out I had to get a job. I was always curious about filmmaking, so I started driving vans and sweeping floors for production companies. Eventually I was coordinating, then producing and directing because degrees weren’t really important in that world. As personal computers and new digital cameras transformed the industry, I found myself creating media for the internet and building digital production facilities for online commercial content. It was all new and there were no real experts; I just kept walking into the rooms where these things were happening and adapting. Eventually I became a product director building AI applications.

Along the way I developed a flawed ability to predict the advance of technology. I told my colleagues film was dead. I said YouTube was going to be the largest network, ever. I was one of the first non-college emails on Facebook and blathered how social media would kill magazines and newspapers. Of course we know what happened, but there was an interesting offset to all my predictions - everything always happened faster than I thought.

The Beginning of AI

The first distributions of AI available to build commercial web applications were clunky and fragmented. It was difficult to write and test the calculations to connect a speech-to-text API to other AI engines that could extract concepts, analyze sentiment, or use visual recognition to tag images. There were limits on voice-activated tools on mobile devices; there were processor and network constraints. But we quickly made powerful applications that could learn about you from social media photos, talk to you about your finances, build marketing cohorts across massive data sets, and predict when you might walk into a Walmart and what you might buy.

I saw firsthand how fast this was going to move.

I knew that computers were really good at learning patterns and executing routines. And I saw how many jobs in the global agency where I worked were just routines. Complex routines, but still. In finance, HR, project management, media buying, data analytics.

When a colleague heard my concerns he joked with me about the coming of Skynet. I remember saying it wouldn’t be like that; there would be no centralized monster machine. It would be ubiquitous applications that amplify and replace the productivity of everyone in the office. I could see departments of 100 people, like finance, being reduced to ten workers or less. I told my friend it would happen really quickly and the job displacement would be outside the bounds of anything we’d ever read about.

And there would be two types of people governing the crisis: aging, self-obstructing politicians that can’t understand email, and technology titans with a favorable view of eugenics who would see the unemployed as excess capacity.

It took ten years.

Most estimates say anywhere from 30%-50% of knowledge-worker jobs will be automated by 2030. That’s only five years away.

chart: world economic forum

The dark purple in the chart above are task areas expected to be fully automated. This is a map for layoffs across industries. And note the top categories: banking, insurance, software, capital markets, energy. Those industries are typically invulnerable to volatile disruption - no more.

In 1999 automation made a huge number of jobs go away. Some never came back and the new ones took a decade to arrive.

Today’s global uncertainty is driving companies to shed labor and AI automation is going to keep them from hiring it back. This time the exodus of employment will be larger and more sustained.

What to do now

Recognize that there’s a rising chance you’re going to have a gap in your employment, with the likelihood increasing over the next five years.

A sincere word of advice: don’t personalize it. This is not being done to you, it’s happening to everyone around the world.

If you’re an individual, the best course of action is to focus on how to adapt.

If you own or manage a company, focus on how to organize around these tools quickly.

Don’t Believe the Un-Hype

Skeptics will point out AI’s flaws as a reason to dismiss it. They will tell you AIs are going to fail because they lose accuracy when they run out of human training data. I remember people who insisted readers would always buy the paper edition of the New York Times because, “how do you hold a PC in your lap while relaxing on the couch?” Sarcasm didn’t save those jobs.

Don’t listen to anyone who believes technology stops because people don’t like change.

The challenge of AI hallucinations will be solved, probably by the AIs themselves, and they will continue to improve. In 10 years AI will surround you, in every aspect of your life. (Remember my record for predicting timing.)

Use it

Start using AI, immediately. Especially if you don’t think you have a use for it. McKinsey and PWC estimate that the effective roll out of AI tools in companies has barely started and that only ~20% of employees are effectively integrating AI in their workflow. If you can show you’re already using tools to help increase your output you’ll have a better chance at continuing to grow with your company.

I put a list of resources down at the bottom of this article. Use any of them for 10 minutes a day. Be regular about it.

Pro tip: ask the AI for its sources and check the answers. Just like a web search, sometimes AIs answers can be wrong. Demonstrating you know how to validate them efficiently and improve results will be a useful skill in the near term.

Pro Tip: tell your boss you want to help evaluate the tools acquired for your department. 90% of AI models available now can only answer direct requests; they operate passively, driven by human input. Next up is Agentic AI. It will be proactive, able to automatically take on and complete individual tasks or entire department workflows. Learn everything you can about current tools so you’re ready to work with the new ones as they’re deployed.

Pro Tip: don’t do this in secret. Share your questions and learnings with your bosses and co-workers. Discuss how you can use these tools to be more productive in ways that are ethical and compliant within your company. Being the person who asks these questions puts you on a different shelf.

Chart source: Axios

35-55: When You Lose Your Job

Don’t look for another one. Look for a different one.

The automation wave after 1999 saw new college graduates and junior-level employees out of work for an average of 6-12 months. For experienced workers it was even longer, 6-18 months.

People in heavily automated industries were hit the worst. Direct mail marketing, print advertising, chemical photography, or editorial content for magazines. Those people eventually had to transition to other careers, often taking lower wages to start over.

Be candid with yourself if your job in the finance department, law firm, communications agency, or tech company is coming back and consider changing the focus of your career.

Pro Tip: your job is not your identity. I have spent my entire career being automated out of my skillset. Film died, I mastered digital video. Post production moved onto computers, I built a digital editing system from stock parts and added that skill to my resume. Distribution went from cable/DVD to online streaming, I taught myself enough code to manage developers and build commercial streaming platforms. I’d be a fool to start that career now, but new spaces are always opening.

There will be growth in:

• Companies specializing in nutrition or memory care for an aging population

• Lobbying for and building high-density residential units for affordable housing

• Providing new forms of community and home-based schooling for tax-strapped cities and towns

Being open minded can be a superpower.

Gen-Z & Millennials: Run for Office

I mean it.

It seems like Millennials in particular were the generation the US was out to screw over. They entered the workforce during the dotcom bust, were hit by the banking crisis in their 30s, inherited impossibly high housing costs and the employment crisis of the pandemic, now this.

AI is going to create a massive displacement of human beings that should have every opportunity to transition into new situations where they can thrive. Corporations are not built to carry this burden and no amount of “compassionate branding” is going to change what the market compels them to do.

The tools that solve this problem aren’t driven by capital markets, but by civic engagement.

Pro Tip: if you believe government should be responsive to the needs of your generation, then it’s time to be the government.

Our Senate, House and Executive Branch average over 60 years old, with leaders mostly over 70. It may be ageist to say this, but we need people who actually understand how technology works and the immediate impact it has on lives.

So if you find yourself out of a job, know that it is totally possible to take theirs.

Pro Tip: 1 out of 20 elections in Congress had only one choice on the ballot.

Pro Tip: In your State elections, the number of districts with no competition averages 70%.

The best estimates available show 18-35 year-olds average from 1000-1600 connections on social media. Your local state rep or senator averages far less than that and they usually have to pay serious $$ to buy email lists for their campaigns. You’re a social media native, and you have a no-cost network advantage.

Build your message, build your brand, and build your political career. I’m with you.

Managers: When You Lay People Off

Motivation is going to be key. The most important thing for people to stay engaged and believing in your company is having a vision for how you’re going to adapt.

Conduct exercises across your company to interview people about AI, how they see it changing the world and how they think it can, or should, change the way they work. Turn every single one of these interviews into a credible idea for how your company can use a new service, create a more efficient process, or build your own AI applications. Organize these according to business group, then plot them according to difficulty and urgency. Start executing your top 3.

A great colleague I used to work with said, “The client expects us to be the expert that’s there to help them, so we have to act like it.”

My experience growing and leading teams that built AI applications taught me that the people you need are mostly in your company already. It’s not obvious because AI is so new no one feels qualified to say they can work on it. But in reality everyone has to be, and some of them will to turn out to be really talented.

Realign people who are enthusiastic about change into work groups where responsibilities and goals are prioritized over their particular title or experience.

Encourage them to remember these tools exist to improve work, and work should exist to improve quality of life. The best companies will strike the right balance and thrive.

Anyone: Start a Company

Your skills are still useful outside a large company structure and now you have a whole team of digital assistants ready to work for almost free. The services you provided inside a corporation are needed by small businesses all around you who could never afford them before. Skills in billing, hiring, customer service, healthcare procurement, and marketing. Use AI to deliver services small companies need to grow.

Organize

We’re living at the apex of a huge change that started in the 1970s, of technological growth and wealth concentration. Techno-capitalism has brought us a lot of amazing things and generated a ton of money, but it hasn’t been good at spreading it around.

A couple events in first part of the 20th century offer some important lessons.

After the First World War when the stock market collapsed and caused the great depression, the US government was unable to pay benefits promised to soldiers who fought overseas. With no hope of employment, veterans traveled from around the country to Washington and protested. President Hoover, a friend to the robber barons of the gilded age, ordered the current army to violently beat, and even fire upon their former brothers-in-arms to disperse them.

2 decades later, after WWII, the US government did something completely different: the G.I. Bill. It provided a basic income, housing assistance, and free college education to active duty veterans returning from the war. What happened as a result? The greatest explosion of highly trained workers in the history of the planet who went on to build the companies that created the world’s most powerful economy. The space program. The Interstate Highway system. Commercial jet travel. Television. The beginning of the computers that created the Internet.

And what happened as the wealth of this new “middle” class was generated? The 40-hour work week. The Civil Right Act. The Voting Rights Act. The right for women to have their own bank accounts and credit cards. The Gay Rights movement.

Wealth brings freedom.

AI is going to create incredible wealth. But the proper distribution of that wealth is the key to our survival and future growth. It won’t just happen on its own, but it has happened in the past and we can work to make sure it happens this time. Ask your state and city representatives about AI. Organize locally and learn about UBI.

We have a once in a generation chance to use this moment to invest in the people this technology was meant to serve. If we do it right, imagine the things we can accomplish with it in the next century.

Tools to use 10 minutes a day

It doesn’t matter what you do at first. Ask for recipes. Ask it to tell you the constellations. Make it pretend to be a person. Ask it how it can help you in your job. Ask it to put data into spreadsheets with formulas. Tell it to build you a website, then ask it to explain the code in terms you can understand. Play with pictures. Watch videos about LLMs and diffusion models that explain how these things work.

For search, text, code, & data: ChatGPT • Claude • Google Gemini • Google Notebook LLM • DeepSeek • Qwen • LeChat • Phind • Perplexity • Co-Pilot

For spreadsheets: Formulabot • Sheety • TheBricks

For CRM: Sheetsy • Agentforce

For images: Stable Diffusion • Leonardo AI • Ideogram AI • Recraft AI • Freepik • Krea AI • Craiyon • Adobe Firefly • Canva

For video: Sora • Runway • Invideo • Descript • Veedio

If you have a company that has challenges in this field, feel free to reach out.

Disclosure:

I used different AIs to assist with the research for this article and personally vetted and fact checked the results. AI did not write the words, I did. So, sorry for the bad grammar, overuse of commas, and occasional typo.