8 Things Two Years in AI Taught Me That Nobody Posts About

AI is threatening everyone. That is the feeling, anyway.
I’ve been working in AI for about two years now. Most recently that’s been at Arcade.dev, helping teams get their agents into real production use. I’m still figuring a lot of it out, like everyone else.
Over these two years, the same questions keep finding me. A friend over dinner wondering if their job is safe. A customer trying to figure out where to even start. Someone new to tech, and someone with decades in it, both quietly asking what actually matters now. I don’t have all the answers, but I’ve collected a few honest ones. Here are eight things I keep coming back to.
1. Do not doubt the skills you already have
What you have built over years is legitimately useful, even if the ground is shifting under it. Your old skill may be transcending into something else. Figure out what that something else is. But do not start by throwing away what you have.
Compounding still works in skills. It is still working.
And you’re not the only one feeling it. Everyone is adjusting to how fast things are moving. Everyone is going through a disruption in their career and their life, and everyone is working hard to get through it. It is not the whole world ending. It is the whole world changing at the same time.
And if you’re just starting out, you have something the rest of us don’t: no baggage. You haven’t spent ten years learning “the right way” to do things that may not matter anymore. Use that. Try the weird idea. Just remember that a fresh idea only counts once you’ve built it.
2. Meet people in person
The digital world is getting weird. More automation, more bots, more noise. So do more of what AI cannot do.
Have food with people. No robot is going to split a pizza with you and ask how your family is doing. Go meet a friend in person and say hello. It opens up perspectives you will never get inside the tech bubble.
That is how I got my job. I met people in person. A LinkedIn profile is a vastly different thing from the person standing in front of you.
3. Your originality is the leverage
There is exactly one person in the world who is like you. No one can replace that. So go after it, big time. I know it is hard. Go after it anyway.
If someone comes after your originality, they have already lost, because they are admitting they do not have their own.
There is always some value you can add in a given situation, regardless of money, title, or anything else. And that value compounds too. Test it.
4. Build something real. Not something fancy.
I see a lot of people talk about agents. Some are even trying to build critical agent infrastructure, and they have not built two agents for themselves.
Building agents is hard. It is not running a prompt. There is deployment, scaling, quality, evals, design, model choice, prompting, cost. Do not underestimate any of it.
If you’re new, this is your edge. Almost every hiring manager will ask the same question: what have you built? So build. Contribute to an open source project. Ship a rough idea. Put something out there that works, even if it’s ugly.
But don’t confuse building with building a lot. Just because you can build something doesn’t mean you should. Knowing what not to build is a skill too. Building what people actually need is vastly different from building something fancy that no one is going to use.
Talk is cheap. Build the right thing.
5. Adoption is slow. Wait for it.
Almost none of this technology has reached real scale yet, outside a few model companies. True adoption happens slowly, because habits are hard to break.
At the end of the day, humans are the consumers of this technology. They pay from their own pockets. They are the economy. AI will take time to make money, and that is exactly where you can make money too, if you are in the industry.
So do not conclude too early. Do not get frustrated. Do not build FOMO. I am not an AI optimist. I am skeptical of many things. But doomsday is not coming either. There will be a place for you to keep adding value.
Wait for habits to break. Wait for people to embrace agents. Wait for people to understand the patterns.
6. Join problem solvers, and ask hard questions
In any industry, look for the people solving a really hard problem. Find that group and join it.
Do not be fooled by a fancy marketing site. In my experience, the people who do not know the depth of building a product end up with the fanciest marketing. The people building great products often do not have time to polish theirs.
So when a company vibes well with you, question it deeply. Who are their customers? How will they sustain long term, financially? How do they make money? If the market crashes, do they have enough funding secured to survive it? What are their case studies? Do they have a sales team?
Ask those questions before you join, not after.
7. Study the past
History repeats itself. That is why we call something a pattern: because it is one.
You have all this fancy AI. Use some of it to study the past. What happened when the internet arrived? When cars were invented? When world wars reshaped everything? You will find clues about what is going to happen to you. Use them as a wedge to reinvent yourself. There is no other choice.
8. Nobody retires from what they love
Billionaires will tell you money will become invalid and nobody will need to work. Regardless of all that, your hard work, your focus, and your perspective are not going away. Your energy is always going to be there. You have to put it on something.
Do not fall for the retire-at-forty trap. Nobody actually retired. Some people made more money. Some stopped working for a company. But nobody ever retired from their passion. Nobody retired from doing the thing they love.
Working for yourself, building something meaningful, that is not going away.
Do something meaningful.
So, who are you going to meet next week? And what agent are you going to build tomorrow?