And lots of upselling at the gym, sports water subscription, energy bars

That is awesome, congratulations! Becoming a dad was for me the biggest but the best change that happened in my life.

@birchtree.me Made an interesting video on this, basically you would need raindrop or Readwise to capture the links and send them to Obsidian. Than publish from Obsidian to your blog

https://youtu.be/JMuzAE2HC20?si=YYLEtrNnw020ex1x

Well deserved!

I didn't get a notification from your reply in the blog post unfortunately. I replied using the Bluesky comment button.

Awesome to hear about your car options, the Enyaq and Nio 5 are also very solid options!

Dit is overigens precies het probleem in veel debatten, er wordt te veel geroepen zonder raadplegen van feiten en onderbouwing.

Het zou absoluut de toekomst zijn als AI Agents een rol gaan spelen in debatten al betwijfel ik of we op dat punt zijn gezien de foutgevoeligheid van veel LLM's.

Bonus points if the field does not validate when you paste in a value

True, but model selection directly impacts architecture decisions. Different models have varying resource needs, latency profiles, and scaling characteristics that shape how we design the supporting infrastructure. They're interconnected challenges.

I agree it’s simpler on a small scale. But at the enterprise level, building an AI Service Agent means tackling challenges like implementing strong security protocols, managing complex data flows, and ensuring compliance with regulations.

Have fun!

Fun fact, this seems to be the first #agentforce post on Bluesky, hopefully many more will follow!

▶️ Technical Performance Metrics:
- Model size and parameter count
- Contextual understanding depth
- Inference speed
- Latency performance
- Multilingual support
- API accessibility
- Licensing cost structures

▶️ Computational Considerations:
- Resource requirements
- Infrastructure compatibility
- Scalability potential
- Cost of deployment and maintenance