Timely Joe. I was just explaining this to my boss. The build vs. buy math equation just became unhinged, but it will be super hard to "vibe the enterprise" because all of the sheriffs are so risk-averse. SMB's for the most part won't vibe, I can't imagine a 20-person plumbing contractor doing that. And big enterprises won't vibe. However, it's not the "I'm gonna be cloned" threat to the giant seat-license app providers; it is the fact that their seat count has to shrink in an AI world.
I’m shocked too. Seems like everyone is using AI for nearly everything at work. Especially prevalent is shadow AI if employers can’t provide decent access. It’s everywhere
My current dream is not that systems of record go away. There is good reason SAP is as complex as it is, it’s the heart of many big businesses but that I can do my workflow any way I want and have MCP and Agent2agent manage the go between.
I have had AI build the wireframe for the front end of my farm website/app.
My coding background is in python and I don’t have any first hand experience with Java or react. I have used Claude exclusively ($20 tier), and besides the constant changing conversations, ithas done a bang up job. I still of course oversee everything it spits out and I ask for detailed explanations of each of the code blurbs. Plus I can see changes live by running locally and we are able to iterate through what stays and what sucks.
I fully understand it’s not perfect and I catch stuff all the time. But it created something of value using my Ideas and thoughts. And it did so significantly faster than I could. I would say even average coders (like me) can be incredibly dangerous/productive as long as you are intently reviewing the outputs, and providing AI adequate guard rails to your problem.
I fully agree ai is rapidly mastering coding and the speed at which it codes make its incredibly formidable in the right hands.
> Valid concerns. Absolutely. But do you really believe the AI labs are unaware of these issues? Be very, very careful extrapolating the shortcomings of today’s models to what’s likely sitting in the lab right now, 2 to 5 years ahead of what you’re using. The teams at Anthropic, OpenAI, Google, and others aren’t sitting around ignoring hallucinations. They’re engineering solutions. They’re building better architectures. They’re creating more reliable systems. Tomorrow will look vastly different from today.
New models need more compute for training. "Reasoning" needs more compute during inference. And what happens when the funding stops?
Spot on. Could I AI-generate my next Pilates routine?
Timely Joe. I was just explaining this to my boss. The build vs. buy math equation just became unhinged, but it will be super hard to "vibe the enterprise" because all of the sheriffs are so risk-averse. SMB's for the most part won't vibe, I can't imagine a 20-person plumbing contractor doing that. And big enterprises won't vibe. However, it's not the "I'm gonna be cloned" threat to the giant seat-license app providers; it is the fact that their seat count has to shrink in an AI world.
I’ll have to disagree based on what I’m seeing right now. Seeing the biggest gains from nontechnical people vibing new workflows, especially in smb
What kind of SMB? Technical makes sense. Non-tech I’m not seeing it.
Law firms, gyms, doctors to name a few. Using ChatGPT, Claude, etc
Wow. OK it’s happening faster than I thought.
I’m shocked too. Seems like everyone is using AI for nearly everything at work. Especially prevalent is shadow AI if employers can’t provide decent access. It’s everywhere
Well the motivation is that every hates “clunk” and their boss doesn’t care, so they fix it themselves.
My current dream is not that systems of record go away. There is good reason SAP is as complex as it is, it’s the heart of many big businesses but that I can do my workflow any way I want and have MCP and Agent2agent manage the go between.
I have had AI build the wireframe for the front end of my farm website/app.
My coding background is in python and I don’t have any first hand experience with Java or react. I have used Claude exclusively ($20 tier), and besides the constant changing conversations, ithas done a bang up job. I still of course oversee everything it spits out and I ask for detailed explanations of each of the code blurbs. Plus I can see changes live by running locally and we are able to iterate through what stays and what sucks.
I fully understand it’s not perfect and I catch stuff all the time. But it created something of value using my Ideas and thoughts. And it did so significantly faster than I could. I would say even average coders (like me) can be incredibly dangerous/productive as long as you are intently reviewing the outputs, and providing AI adequate guard rails to your problem.
I fully agree ai is rapidly mastering coding and the speed at which it codes make its incredibly formidable in the right hands.
That’s super cool Sloan!
> Valid concerns. Absolutely. But do you really believe the AI labs are unaware of these issues? Be very, very careful extrapolating the shortcomings of today’s models to what’s likely sitting in the lab right now, 2 to 5 years ahead of what you’re using. The teams at Anthropic, OpenAI, Google, and others aren’t sitting around ignoring hallucinations. They’re engineering solutions. They’re building better architectures. They’re creating more reliable systems. Tomorrow will look vastly different from today.
New models need more compute for training. "Reasoning" needs more compute during inference. And what happens when the funding stops?
As one example, look at the work being done on small language models