All true, but a fluctuation of a degree or 2 in the routine demo could put a plane in the crowd
I woulda loved to been there too, also think it’s cool as hell,
I think the camera angles + the aircraft size makes it look worse than absolute reality but definitely pretty close
Never flew high performance aircraft and never been military, I definitely understand the expert opinions and completely respect them on both sides
My personal opinion is not make an example but set some new guidelines, have experts in the field set the parameters and make the pilots follow them
That being said the Blue Angles and their leadership should be the subject matter experts in this case, they perform a function that is outside of normal fleet flying and are meticulously selected. They should be their own judge and jury, hell they critique every single performance on themselves who better to decide?
As much as you enjoy these builds you should try your hand at building and flying RC airplanes
It’s super fun and the people at the local clubs are great and usually incredibly helpful when learning
My dad got into it in retirement and I think it was one of things he enjoyed most as a hobby
If you really want a challenge try the helicopters w the real cyclic and collective, the ones without all the auto stability, I was always a fixed wing guy till I got hooked on the RC helos, demanding and extremely satisfying to be able to control one.
Yes, and I saw some local news footage that proves it’s accurate. Could an onshore breeze help the wake turbulence carry onto the shore more than normal?
I think it’s cool as hell and wish I was there, but I can definitely understand and always respect some of the MUCH more experienced opinions of those that have actually flown high performance aircraft. And about the rules/discipline aspect.
I just wonder when I see these things about the camera angles and perception vs reality, the Apache incident looked much worse from one angle than another.
While it’s no tennis court like the 14 those E models are pretty good size jets.
@RSE_VB When you see it from the balcony view I don’t think he was even over the sand. I think the size of that jet especially when compared to the A model creates the perception it’s allot closer than reality.
Definitely close but directly overhead.
The Blue Angels are conducting a “thorough safety review” this morning after circle and arrival maneuvers yesterday at Pensacola Beach, in which “an aircraft flew lower than standard profiles, resulting in a disturbance on the beach that affected civilian chairs and umbrellas.”
@GeorgiaMan45@BlueAngels@DeptofWar It’s never too low if you can still fit another blue and gold plane between it and the ground. These guys probably could fit 2 w room to spare.
Let’m app up the home crowd, looked like everyone lived it to me.
@bcherny Am I missing something or is what I am trying actually working?
No Claude.md, run Fable to plan a multi step change, max effort for plan.
Review plan, everything ok,
Revise plan on last prompt, “use lower tier subagents to perform all tasked based on complexity, this task does not require max reasoning so conserve tokens by assigning agents most suitable for the task”
Been running for a while now on only pro plan and seems like less token usage than sonnet 4.6
Does Claude actually understand use lower tier of asked explicitly?
@RSE_VB@wbridgefa Split rim tires/wheels can be very dangerous even at relatively low pressures, my dad had a coworker at Grumman killed by one. Tractor trailers heavy equipment and aircraft tires need to be broke down/replaced carefully and usually inflated inside a safety cage.
@bcherny Claude code idea
When a user starts a task and runs out of a 5 hour limit, option to pick up automatically on 5 hour reset.
Add switch to context window and add auto global like send all to remote.
This way tasks can complete without user needing to manually resume if desired,
@domasbitvinskas@bcherny Started for me when I tried to delegate subagents for simpler tasks. Instead of putting it in /.claude.md I prompted directly with sub agent instructions, then it started eating my paid usage. Never seen this behavior before.
@bcherny I think there may be something wrong w the way Claude is calculating usage. My 5 hour was at 79% and it is currently devouring my usage credit paid.
Is this correct or should the 5 hour run out before hitting the paid.
@PeteHegseth
What do we have to gain by restricting the US public from an AI model?
What do we have to gain by restricting the US govt and its defense contractors from the best AI models?
Do you think our adversaries are not going to buy trillions of tokens and not use the advantage against us?
Or develop their own??
Do you want to drive frontier model creation outside the US??? Set us back years due to a philosophical disagreement?
You are playing a dangerous game of chicken and letting your ego override your brain.
You are going to hurt the very thing you think you are protecting.
@elonmusk
So, what do you think about the latest Anthropic restraint by the US gov?
You have worked private and in the bowls of the machine/swamp, your opinion would be greatly appreciated.
Does this apply to new models released by other companies?
@AnthropicAI This should go all the way to the supreme court.
This is a handful of individuals in the US govt that are acting like babies and trying to cripple anthropic over a philosophical dispute.
Pathetic that those running the US govt can't act like adults.