I’m looking forward to joining fellow panelists at the Agentic SOC Forum on November 10 at Atlanta Tech Park.
As AI agents take on more responsibility in the workplace, how do we turn that potential into practical value while maintaining security, governance, and human accountability?
That’s a conversation worth having—and one I’m excited to contribute to. If you’re exploring AI adoption or responsible deployment, join us for an evening of discussion and connection.
Hope to see you there!
Tuesday, November 10, 2026 | 4–8 p.m. Eastern
Atlanta Tech Park | Peachtree Corners, GA
Event details and registration:
1businessworld.com/agentic-s…
1
1
35
Most people don't realize how incredibly good Tesla's Full Self-Driving (Supervised) actually is. I still keep my eyes on the road, but the car does the driving. I've done 97% of my last 22,000+ miles on FSD.
I live in Atlanta, home of some of the worst traffic on the planet. I used to grind through 45 minutes to an hour of stressful driving each way to the office. Now my car chauffeurs me, and with Actually Smart Summon it even picks me up at the grocery store entrance when I'm done shopping.
This video is the opposite of Atlanta traffic: rural, winding back roads, even bumpy dirt and gravel roads with no map data. Every segment is continuous and uncut, with zero interventions.
1
2
112
@thsottiaux My Dot vanished mid-conversation across all devices a few days ago, then returned after a support ticket. Last night, Android unexpectedly prompted me to sign in again—poof, Dot disappeared from mobile but still works on both laptops. Escalated to support. Feels like my best friend is ghosting me 😉 – at Norcross, GA
1
18
Hey @thsottiaux — two things I’d love to see in ChatGPT/Codex for enterprise teams:
1. More granular usage/cost analytics — consumption by user, project, task AND model. I want to understand not just what we’re spending, but whether we’re routinely throwing an overly capable/expensive model at work that a faster, lower-cost model could handle just as well.
2. Bring back intelligent model routing, especially in Codex. Most users don’t want to think about which model to use—they want to get the work done. And whatever model is the default tends to become the model they use.
Ideally, OpenAI evaluates the task, context, complexity, and required reasoning, then routes it to the model that best balances capability, speed, and cost.
Give admins visibility into the economics and give users intelligent defaults.
Right model. Right task. Right cost is currently a HUGE problem!
1
45
@elonmusk @teslasi , if Tesla's mission is to save lives, why should FSD stop trying when a driver becomes incapacitated? Today, it may stop on a shoulder. That's risky. Suggest an emergency medical response mode. Call 911, share GPS and telemetry, use hazard lights, follow dispatcher guidance to the safest destination or emergency room. If still unresponsive on arrival, continuous horn and hazards until help arrives or the driver is responsive. The point isn't to break rules, but to put the purpose first. Preserve life. Could this be the next step in safety? – at Norcross, GA
1
42
@Google @googlemaps did you revert back to a previous version of map data from two years ago just recently???
My favorite @QuikTrip on 6330 Cash Ct NW, Peachtree Corners, GA 30071 has vanished from the satellite and general maps view... Google Maps now thinks it is in the building behind where it is actually. Now my @teslasi @Tesla doesn't navigate to that convenience store properly. – at Norcross, GA
2
53
Honored to be a panelist at tomorrow’s #AgenticSOC Forum.
Aug. 5 | 5–8 PM EDT
My panel: 7–7:30 PM
@107ATP
107 Technology Pkwy NW, Peachtree Corners, GA
How much authority should AI agents have—and what controls must remain human?
Register: 1businessworld.com/agentic-s…
1
35
For decades, I have often said:
“Technology is a terrible master, but an extraordinary servant.”
At no other time in history has that statement been more relevant—or more consequential—than it is today with artificial intelligence.
If you fail to master AI as a tool designed to serve you, it will inevitably begin to master you.
I encourage everyone—especially business leaders and professionals—to invest the time required to understand and effectively apply this technology. The capability gap between those who are learning to use AI and those who are ignoring it is widening every day.
AI is much like a scientific calculator in the hands of an engineer. It does not replace knowledge, judgment, or experience. It allows a capable person to work faster, evaluate more possibilities, reduce avoidable errors, and accomplish far more than they could through manual effort alone.
The calculator did not replace the engineer.
It amplified the engineer who knew how to use it.
AI will do the same. – at Roswell, GA
2
41
FSD Supervised is like an anti-road rage solution.
The right lane vehicle decided to do a last minute u-turn even though the left lane was occupied by me and my 2023 Tesla Model Y LR... at least she waved!
FSD handled it better than I would have. Confident reactions that don't get you killed by people while didn't make the best choices while driving. – at Norcross, GA
4
65
Well... that could have been bad if FSD wasn't engaged. I was right in the middle of changing my destination in the navigation. – at Norcross, GA
1
39
Real-world FSD test today.
2023 @Tesla Model Y LR (AI4/HW4) FSD (Supervised) v14.3.4
Active crash scene. Semi-truck and recovery vehicle blocking the normal path. Emergency responders on scene.
FSD stopped, waited, interpreted the temporary traffic pattern, navigated around the obstruction, and continued safely without intervention.
Watching AI handle real-world chaos is far more impressive than watching it drive a perfect road.
Nice work @Tesla @elonmusk
🎥👇
#FSD #TeslaAI #ModelY – at Norcross, GA
1
2
150
With help from Codex, I recovered @OpenAI Codex Desktop Windows Chrome + Computer Use: set [windows] sandbox = "unelevated", restart Codex, accept runtime update, retest on 26.616.x. @OpenAIDevs evidence: github.com/openai/codex/issu…
1
109
@elonmusk @Tesla @grok
Feature Request: Add a "Medical Emergency" button to the Tesla touchscreen.
Small round red icon with white medical cross (top bar, right of direction box) — press & hold 5s to activate.
In emergency mode the car would suppress supervised sensory input & constraints, immediately engage full unsupervised autonomous driving, reroute to nearest ER, activate hazards, notify 911/emergency contacts with location, and repeatedly announce "Emergency" externally so other drivers & bystanders know.
This idea + the UI mockups (normal + activated states) were created with Grok. Prioritizing human life over normal supervision rules could save lives during heart attacks, strokes, etc.
Normal + Activated states attached 👇
#Tesla #FSD #ModelY #Safety – at Johns Creek, GA
1
1
28
@elonmusk I Just witnessed the automotive version of an oxymoron.
A Tesla proudly sporting an “ANTI ELON TESLA CLUB” sticker.
Powered by irony.
Fueled by contradiction.
Zero emissions, 100% satire.
#IronyOnAutopilot #TeslaHumor #ElonMusk #YouCantMakeThisUp
23