It don't mean a thing if it ain’t got that swing
🎵 Jazz, 💰 $2b seed rounds, and 👻 fake job applicants.
I can still remember the smell.
All music instruments have a smell to them but the one that I remember most distinctly is the trombone. The smell of metal and spit with a hint of mold is something that sticks with you for your whole life.
I started playing trombone in 4th grade and by middle school I fell in love with playing jazz. By college I was playing in the jazz band at UNLV —I even had a tiny scholarship.
I'm not talking Kenny G here. In the jazz community there is a set of "jazz standards". These are songs that all jazz musicians play and listen to hundreds if not thousands of times over their lives. The standards are very simple songs that usually fit on one piece of paper and are typically what you would hear a jazz trio playing at a restaurant like Bandera --RIP Bandera West LA.
The layperson will look at something like this and probably think, "how is it that jazz musicians can play the same very simple song thousands of times and still love it"? The answer comes in the form of a song made famous by Duke Ellington: "It Don't Mean a Thing if it ain’t got that swing".
The statement that this song makes is pretty clear in its title. The thing about jazz is that if you just play the notes thats not enough. You need to play the notes with "swing". Swing is hard to describe. Its a feel...a style...an attitude that anyone who plays jazz brings to the music.
Where does swing come from?
My music teachers instilled in me a deep appreciation for the fundamentals. They drilled scales, chord progressions, and music theory until the skills were second nature. But just knowing the instrument isn’t enough for jazz.
For me swing came from growing up with jazz music playing in the background while I played with Legos. It came from humid summers riding my bike around the south side of Chicago and playing baseball with friends. It came from fighting off bullies in the school yard. It is a compression of all of my lived human experiences and thats why it was unique to me.
Swing feels more relevant than ever today as machines increase their grip on our day to day lives.
For all of human history, you could brute force your way to success by being good at the fundamentals. The winning formula was simple. You lock in and outwork everyone else. During college I realized this could be an edge and made it a core part of my identity. I’m reasonably intelligent but being willing and able to brute force any problem has been a major factor that has created the life that I live today.
Mechanization removed the ability to win via brute force with much of the physical work that was done 100 years ago. And now LLMs have come to mechanize knowledge work. Anthropic’s Claude (my model of choice these days) doesn't get tired, it doesn't need breaks, and it doesn't care if the task is repetitive. It just executes.
In the world where Claude exists, sheer effort doesn’t cut it anymore. I realized this in mid 2022 and it sent me into a tailspin. I ended up pretty sad and worried about what I was going to be able to do with the remainder of my working years.
But here’s the thing I’ve realized. AI can’t swing. It isn’t human. It doesn’t live a life, feel emotions, or see the world through a unique lens —at least not yet. What it creates lacks the depth and authenticity that comes from human experiences.
That’s where we still hold the advantage. Whether it’s building a product or telling a story, the real value lies in the perspective only you or I can bring. It’s gotta swing.
And the future belongs to those who can swing. Machines will handle the brute force—the repetitive, scalable, and technically precise tasks—leaving humans free to create, innovate, and resonate.
The Duke was right, it don’t mean a thing if it ain’t got that swing. So I’ve been asking myself: what can I do to add more of my unique style to what I’ve been doing.
This is a reminder that you should delve into this question as well 😉.
🤖 AI
Meta was gaming AI benchmarks, which for some reason isn’t surprising. These days all of the models are really good and any improvements are hard for users to notice. Incentives are a huge factor here. When market cap and intellectual honesty start to diverge I start to get more worried about sustainability. Where are we headed?
Higgsfield made a huge splash on the interwebs for its ability to generatively create some amazing camera motion video from still images. Here is a review from a friend in the production industry: “Takes 5-10 minutes to generate one clip. And, like Runway, it f*&#s up a lot. Definitely getting somewhere but still unusable for commercial.” A lot of these applications demo well because you aren’t actually trying to use them for anything practical. Control remains illusive.
Companies are increasingly facing challenges from fraudulent job seekers utilizing AI tools to create fake identities and resumes. Research from Gartner predicts that by 2028, one in four job candidates may be fake. Its funny that on one side of the labor market you have ghost jobs, and on the other ghost candidates.
Mira Murati's AI startup, Thinking Machines Lab, is targeting a $2 billion seed round. A friend of mine responded to this news by saying “why the f do we bother with PMF (product market fit).” He is right…why do we? The real answer is probably something like “in the startup game, product market fit never mattered”. But I’ll leave that for another post.
💡Steal this idea
MCP servers and external tools allow LLMs to access external systems —tool use is how things like search or operator work in ChatGPT. I wish there was a system that let me create bundles of tools / MCP servers that I could easily attach to agents for specific purposes. Ex a CFO agent would act as the interface between myself and my ERP and other finance related stack. A marketing agent would act as an interface between me and my marketing stack. Etc.
🔗 Cool *hit
Someone is removing Garfield from the Garfield comics to “reveal the existential angst of a certain young Mr. Jon Arbuckle”. This idea swings like a mofo.
🔈What I'm listing to
📚 What I'm reading
AI 2027 is a heavily researched month by month projection of what will happen over the next few years.