AI has got really good at coding. Anyone can create something interesting on a computer. Steve Jobs famously described the computer as a “bicycle for the mind” and said that:

“the best way to learn about computers is to get a computer and tinker with it”.

The same is now true for AI. With just a simple prompt you can explore and create something unique and interesting. This is important, we now have a scratchpad for prototyping and experimenting with ideas. Exploring new concepts makes you think creatively, this puts you in the mindset to for building products, creating art and developing business ideas.

Intelligence Cambrian Explosion

Sam Altman says that the cost of using AI will drop by 10x per year, “whilst socioeconomic value of linearly increasing intelligence is super-exponential in nature” (Three observations). This means it’s getting cheaper and cleverer at a rate we haven’t fully grasped.

All of a sudden unprecidented creative skills and opportunities are available to everyone:

• Rapid prototyping of ideas
• Spark creative thinking in areas outside of your expertise
• Evolve and iterate on concepts to build extremely complex outputs
• Deeply research topics and make connections between seemingly unrelated ideas
• Experiment and research designs and styles
• Augment your own skills and knowledge
• Construct personal guides and tutors, to learn faster
• Build new forms of entertainment and art
• Make people see things differently

AI will be like the birth of television

In the 70s and 80s television was one of the most influential forms of entertainment. Kids were often told that they would get “square eyes”, for watching too long. Of course none of that generation has square eyes, but the impact of television made them see the world differently. That generation became much more sensitive to what made good entertainment, overtime principles developed for making great shows and films. People became “famous”, and “celebrities” emerged in a much more accessible way. Anyone who put the work in to act, could become a star.

This will be true for AI too. Kids are in their bedrooms “vibe coding” into existance new apps and tools, that previously were out of reach to all but the top companies with the budgets to build it. These kids (and AI savvy enthusiasts) are untainted by past experience, their minds have the joy of youth and that naivety becomes an asset. See the 17-year-old who created an “AI Wrapper App” to calculate your calorie intake - he can’t code, and yet is making $12m from an AI powered nutrition app he built with AI.

Rise of the Agent Engineers

The drop in price and increase in intelligence has made it coding a commodity. However it is still hard to make these tools work well, the intricacies of “prompt engineering” are still a black art. We are beginning to see the rise of Agent Engineers, people who are experts at making AI agents work well, but also have the software experience to hop across programming languages, tools and cloud services.

Engineering principles and deep undstanding of tools and languages are not going away anytime soon. Even if the machines can “build all the things”, there will still be a need for engineers to hook it all up together, and plan out and build resiliant systems. It is still a complex task building systems that work well, even if one is orchestrating the agents to do the work.

We are now beginning to see the rapid development of AI products, and this has become available to everyone, cheaply. It’s inspiring to see what people create. Creation is a fundamental human activity and makes us see the world differently. The things we build have an emotional impact, they lead to other ideas, triggering both art and science. The continual process of creation makes us rethink what is possible, and leads to greater understanding.

AI is Still Wonky

With all this excitment and hype it’s important to note that AI is not a panacea, the outputs are seldom if ever perfect - there is “wonkyness” in the outputs. This is ok. At this stage in the AI explosion one has to slightly shift ones mindset and tolerate the imperfections. Look through the dirty window to see the potential on the other side. Like Alice in the looking glass, we need to sip the tonic and then a new world opens up to us.

What does this mean for the future?

People building software is accelerating rapidly, we are seeing more and more creative applications of AI every day. Ideas for businesses are springing up everywhere, but this brings a new set of challenges. Ideas can instantly be copied and shared, old moats are eroding. Old encombants are struggling to keep up with the pace of change, and that “motivated kid in his bedroom”.

What does it mean when anyone can create their own tools and applications, without being tied to existing SaaS companies? Will people still purchase services and software when they can create their own? What are the moats that will protect businesses, and how must we adapt to keep ourselves competitive? We don’t know the answers to all these questions, but it appears that there are still some fundamental principles that will apply:

• make something people want
• be the trusted supplier in your space
• build your personal brand about trust and transparency
• constantly refine and adapt
• customers will always seek options that make their life easier
• hard to get data becomes a new moat
• customers will get more and more overwhelmed by choice, so having a clear voice and providing value will always be important
• make original and captivating things

One Prompt Micro Machines

Anthropic has recently released their Claude 3.7 Sonnet model, and it makes coding accessible to everyone. The “Citizen Dev” can take an idea and turn it into a workable demo. Other platform companies will utilise it for their own products, such as Cursor, Replit, v0.dev and many more.

The following was generated with 1 prompt. It impressed me, partly because I just had an idea and wanted to “vibe code” it into existance.

Create a black canvas with many densely packed tiny complex machines. Each machine is a unique white wire frame running some kind of animation. 
Think about the different mechanical smooth animations the machines can run.
Make these complex machines have a very simple style that is beautiful line art, with white and subtle use of greys. 
Ensure that the output is centered correctly, the machines are aligned in a grid, and that the animations move smoothly, with no jerkiness.

This is what it produced using Claude 3.7 Sonnet (high thinking mode):

Go Build…

There are no excuses, anyone can bring simple ideas to life. Go build your micro machines… and shift your thinking to how you must adapt in the age of intelligence.

Please checkout my AI applications at Agent Web Factory, and follow me on x.com/emson.

Thank you!
Ben