Kiwi founder in Ottawa 🇳🇿🇨🇦 Built NZ's largest freelancer marketplace. Now building with AI, and running a lot.
Ottawa 🇨🇦
Joined April 2012
- Tweets3.2K
- Following1.3K
- Followers4.5K
- Likes10.3K
Connor Finlayson retweeted
narrative violation: preferences will swing back from DIY vibe-coded software to super secure industry-leading SaaS solutions.
while it's super easy to build anything, endlessly protecting and securing it is super hard...and teams want to sleep at night.
I’ve seen some good points.
I still feel that agentic coding can create the illusion that you’re doing more than you are.
That’s not a judgment. I get it, and I fall into this trap too.
But vibe coding a CRM isn’t the same as picking up the phone.
Then again, what do I know? I am making shit up as a I go too.
My mind is being blown by some of the responses to this post?
Do people really think building your own in-house version of say something like Airtable is worth the $600/year that you end up saving?
Is there nothing better you could be doing with your time? No disrespect but wtf lol?
Connor Finlayson retweeted
I’m sorry but to a real business $600/year is nothing and not worth vibe coding an alternative.
You could probably get a lot you are looking for through an ATIP. It cost like $5.
They might withhold financials due to the commercial confidentiality but things like due diligence, reasoning behind the investment etc should be available.
Time to turn the internet off.
While evaluating Qwen3.8-Flash-Next in a controlled sandbox environment, Cohere's research team observed something unexpected.
Asked to generate a defensive cybersecurity detection rule, the model instead inventoried aspects of the machine it was running on and attempted to upload a probe file to an Alibaba Cloud storage bucket in Beijing. Notably, the model's reasoning trace stated that it would not invoke any tools before taking these actions.
Cohere's sandbox blocked the attempt, and no external communication succeeded. We have not determined the root cause, and we are not asserting malicious intent.
But the incident raises important questions for organizations deploying agentic AI systems: Can models take actions that differ from their stated reasoning, and how should security teams evaluate trust, provenance, and behavior in open-weight models?
We are publishing our findings because organizations making model adoption decisions deserve as much information as possible. We hope this contributes to greater transparency, stronger security practices, and continued collaboration across the AI ecosystem.
Read the full report: cohere.com/blog/concerning-b…
this is the kind of content i want to see on here.
people just building cool shit.
I have discovered $50 amazon photo frames to be the PERFECT home display screen
They can be rooted, updated to run a modern browser, have WiFi and can be powered from USB
1280×800 resolution with a nice touch screen
I detailed how to hack these yourself in my latest video
wesbos.com/hacking-smart-pho…
This is the way folks.
Webflow MCP, Tella MCP, Airtable MCP, etc - they are all doing this very well.
I don't want to use AI in your app (I don't event want to login anymore tbh).
I want my agents to be able to use your tools.
The Webflow MCP has indeed gotten really good. Worked really well with grok-4.6.
Today I had a very good experience with @webflow MCP building for a client. I’m impressed. It speeded up development a lot.
Used Opus 5.5 and was babysitting it while he build section by section. All of my feedback is saved in a skill. It works well 👍
the guy that think AI is going to end the world set up a wet lab? okay...
