Joe Talks AI (@JoeTalksAI) header
Joe Talks AI (@JoeTalksAI) profile picture

20+ years writing code, now building AI agents beyond the demos. Engineering leader sharing real-world lessons on voice AI, agent workflows, security and trust.

Joined July 2025
Happy Friday, PSA to always hallucinate responsibly! Haha
1
1
15
This is phenomenal @bot save me phone tree time!
1
1
26
I see posts from friends regarding dealing with bad Voice AI experiences. Sometimes you just gotta bring your own Voice AI to the table! @bot can now make calls for your with your phone number!
35
Muse renewed my domain automatically, on the date and time I chose, and with my approval. Only glitch was having to give it my 2FA code. Felt like I did the first time I wrote a screen scraping script back in 2005 to automate logging into an older system at my job with no API to update data automatically.
1
5
185
Google Argon 4 just dropped for me. Time to see what it can do!!! #AI #Argon #Gemini
2
5
89
Who's gonna tell the guy throwing the beer can in the air, that AI can do this, perhaps faster, better, more controlled? Image to video models have advanced so quickly, that hybrid is a thing. Perhaps not as much at the local level but it's definitely there. #AI #Video
3
46
Good Morning, time to fly for a conference. My @OpenAI Dot organized everything for each day of the conference, including travel logistics, sessions I should attend each day, the hotel, cost analysis for a rental car vs Uber. I gave it all the details last week and then went about my day. Yes, I could have just asked ChatGPT for an "instant" answer but I didn't really need that. Thought popped up and I said you know let Dot do it, "hey go do this and let me know later on." @TheRohanVarma #AI #DOT #OpenAI
🤖 Made with AI
1
28
Man, I keep checking but still no Astra for me. Guess I'm gonna get a good amount of banked resets lol
1
36
I enjoy getting away from my computer on Friday nights. Doing the video tech and taking sideline video for the local high school tackle team.
28
Replying to @bendee983
Independent checks beyond the model may not be an AI Labs recommendation but it's definitely how most companies handle the unpredictable nature of LLMs and Agents.
88
Replying to @unclebobmartin
Was in a session yesterday with an engineer from Meta who was talking about this exact topic. Never let the LLM be the last step between a critical decision and user. Always have then LLM thats coding be behind a deterministic envelope.
2
75