Started using 6.0 a few weeks ago to save some of the quota. Costs more but way burn and actually gets things done.
xitter.cf/boxxa/status/210650243…
Why are so many Codex devs downgrading from GPT-6.1 Sol back to GPT-5.6 Sol right now?
On paper, GPT-6.1 Sol ($2/$10) costs 50% less per token than GPT-5.6 Sol ($4/$20).
In practice, devs switching back to 5.6 Sol are seeing 20%–30% LOWER quota burn and faster fixes.
Why? Because 6.1 Sol was post-trained to over-think every prompt burning 3x–5x more hidden reasoning tokens and running unsolicited test loops on simple 5-line edits.
Have you switched back to 5.6 Sol, or are you sticking with 6.1?
I love that this is every AIs response to how to make money.
Niche AI automation for local businesses (salons, trades, clinics). Install no-code booking agents or lead responders using n8n + Claude. One person + bots handle setup, delivery, support. Land 4 clients at $2.5k/mo via cold outreach. Near-zero capital, 80%+ margins. Scales to $10k without ads or team.
I am done too once my subscription expires. It’s worse than ever, burns usage, and has continually gotten dumber.
Been using Claude and Sonnet 5.5 and has been crushing most things Sol couldn’t get past.
Codex has been completely unusable this weekend after @thsottiaux 's usage 'fixes' on Friday. I am a power user and am cancelling all subscriptions after this weekend. Codex blew through 3 banked resets and over 60k credits over 2x accounts. Clearly something is very broken after the last update. Back to Claude...where my money goes 10x further
People who complain are PvP players who only were able to kill people who weren’t equipped for PvP. Good job you come in with bobcats to kill people with an anvil and hull cracker. They don’t realize some people play the game for trials and the leaderboard that isn’t kill based.
Now there is a PvP mode, they are running into always battle ready players and getting murdered. The enjoyment of easy kills is over and now their game is ruined.
Someone missed a few physics classes.
People don’t understand how impractical the flying Roadster is. If another car gets caught under the thrusters, it would experience the same force as the Roadster’s weight. That’s why they had to flip during this maneuver, to protect other cars from the thrusters.
Readers added context they thought people might want to know
Thrusters eject gas with momentum equal to lift (Newton's 3rd), but the plume expands and dissipates like helicopter downwash. Cars below get spread pressure, not full weight force. Flip is a stunt demo, not needed to protect cars underneath.
x.com/PalmerLuckey/s…
x.com/Swole_Medic/st…
pilotswhoaskwhy.com/2024/01/29/und…
boxxa retweeted
Sometimes a demo is simply to remind the world that your competition has no desire to dream -- and no ability to evolve.
Love this but is this a cybertruck or model 3 that’s gonna be years to see it?
Configurations
– Performance: road ready, seats 2
Active aerodynamics & multi-mode suction system for next-level performance on & off the track
– Thrust: race ready, seats 2
Multi-mode suction system paired w/ rear-mounted cold-gas thrusters that enable 0–60 in <1 second
– Rocket: air ready, seats 1
Multi-mode suction system plus perimeter-mounted cold-gas thrusters supporting forward & upward acceleration plus lateral rotational movement. So...yes, it can fly
Range >300 miles
Top speed up to >230mph (Rocket)
0-60
Performance (<1.6)
Thrust (<0.9)
Rocket (<0.59)
Reserve now –
Tesla.com/Roadster
