Jeffrey Behnke 💻
@jeff_behnke_
Followers
689
Following
6K
Media
350
Statuses
5K
Writing code 20+ years. I build enterprise SaaS applications & AI Tech. 🧑💻 My passion is Research and Development which is why I have 100 projects going 😉
Wisconsin
Joined February 2019
I have a lot of projects going, all of which are planned to be launched. So this is a thread of my current projects. @DevShedIO @Z3Chat @AISpawn124739 @BrandFastPro @SaaSFastPro and many more.
2
0
3
Now this I actually agree with and have said myself. It’s either this or the models naturally degrade under load.
I think model companies are cheating. VC-subsidized tokens will eventually end. There's a huge pressure on these companies to make money. Word on the street is that these companies are not incentivized to offer you the *best* intelligence, but instead give you *cheap*.
0
0
0
This argument is all sales pitch. If you use docker then it should take the same amount of time to setup dev on any machine and OS. That’s the point of docker 😂.
We moved to Linux as the default for new dev hires at 37signals a while back. Our setup flow complexity collapsed. It's just so much easier to automate everything on top of Linux than on Mac. And for our work, Linux is also just way faster since we use Docker for DBs etc.
8
0
9
I think the current product / project I am working on is so big it will only be affordable for enterprise application development. Something that current AI struggles with. #technology #AI #enterprise #SaaS.
0
0
1
I’ve known this forever but now with the microplastics warnings people are looking everywhere for plastic. It’s a little funny that this was not common knowledge.
“Oh I avoid plastics, this is why I drink from a can”. You mean the plastic bag in your can?.
0
0
0
Claude just got lazy and wanted to take the simple way out deleting all of the complex code. "This system is too complex". LOL, of coarse it is, this is what real apps look like, complex systems.
0
0
0
All of this knowledge combined to develop something truly special. Claude gave me an estimate 😂.12-month development timeline.6-8 person development team.$1.3M total budget.Target: 60-80% reduction in development time. All are much more except the last one I believe is low.
I am cooking something right now from everything I have learned over the last 20 years including all the projects that I have developed over the last 6 months. This is going to be an incredible system like no other!. #AI #technology #entrepreneurs #buildfast.
0
0
1
I am cooking something right now from everything I have learned over the last 20 years including all the projects that I have developed over the last 6 months. This is going to be an incredible system like no other!. #AI #technology #entrepreneurs #buildfast.
0
0
2
It's easy to brag about your employees using your product internally when there is NO token limit or cost to the employee.
0
0
0
AI can help businesses automate repetitive tasks, improve customer service, and increase productivity. Don't fall behind - start exploring AI solutions today. Lets talk if you want to know more. #AI #productivitytips.
0
0
0
Models on @GroqInc are so fast it is insane!. Using llama 4 Scout-17b-16e-instruct runs about 20-30 times faster than using Gemini 2.0 Flash from Google. Crazy speedup!.
2
0
6
RT @code: Let's Learn MCP! This new virtual workshop will introduce you to MCP and guide you through creating your first MCP server. Select….
0
166
0
I see absolutely no value in LinkedIn anymore. It’s just a cesspool of spam. Whenever I do open it I have to dedicate an hour to previewing spam messages.
Bro paid money to get around my "don't dm me" rules on LinkedIn just to send this. I hate modern marketing.
0
0
1
I don’t understand why people think because a model can use defined tools and follow instructions that we are close to AGI. We are no where near AGI and this is not it!.
Not enough people realize that Claude Code is really just Claude Agent. It’s a:.- customizable agentic system.- that's *really* good at knowing when & how to use tools .- disguised as a coding tool. And it’s slowly starting to feel like what working with AGI should feel like.
0
0
0
This is exactly what makes us all desire a competitive open source model.
0
0
0
This right here is the best possible answer I have seen so far. We don’t k ow what they are doing to the models, but we absolutely know it is something. When you spend X hours a day with any tool you learn how it responds and when it is responding differently. This goes for.
there's this phenomenon happening with AI coding tools. i'm calling it "subscription fog.". you have users bragging about getting $3k of inference for $200/month. then suddenly things "feel" off. when the same entity controls both:.1. the inference supply.2. the harness
1
0
0
How did they get such an amazing photo of me?.
0
0
0