I started a project this year similar to this with rats. It’s now two axis with tracking and a stereo camera with depth detection. The amount of hours I’ve spent on it is astounding but I’ve learned a lot!
Also, ended up swapping the Pi I started with to a jetson.
It's both lower pressure above the wing (~20% of lift) and the reaction force from pushing air down (give or take the remaining 80% of lift). The main wrong thing is that the air travels faster because it has to travel farther causing the air to accelerate causing the lower pressure that's double plus wrong. It's a weird old misunderstanding that gets repeated over and over because it's a neat connection to attach to the Bernoulli Principal when it's being explained to children.
a classic example of how LLM's mislead people. They don't know right from wrong, they know what they have been trained on. Even with reasoning capabilities
That's one of my biggest hang ups on the LLMs to AGI hype pipeline, no matter how much training and tweaking we throw at them they still don't seem to be able to not fall back to repeating common misconceptions found in their training data. If they're supposed to be PhD level collaborators I would expect better from them.
Not to say they can't be useful tools but they fall into the same basic traps and issues despite our continues attempts to improve them.
How can you create a pocket of 'lower pressure' without deflecting some of the air away? At the end of the day, if the aircraft is moving up, it needs to be throwing something down to counteract gravity.
Exactly. The speed phenomenon (airflow speeding up due to getting sucked into the lower pressure space above the wing) is certainly there, but it's happening because the wing is shaped to deflect air downwards.
The point isn't about how the low pressure is created just that the low pressure is a separate source of lift from the air being pushed down by the bottom of the wing.
No, what still matters (when explaining why the wing is shaped the way it is) is how the low pressure is created. In this case it's being pulled down by the top of the wing.
I’ve used both. Claude more extensively. I’ve had good results with Gemini too, however it seems easier to get stuck in a loop. Happens with Claude too but not quite as frequent.
By loop I mean you tell it no don’t implement this service, look at this file instead and mimic that and instead it does what it did before.
I regularly have Claude Code in a loop where it can't figure out Typescript types and uses unsafe `as` kludges, even when CLAUDE.md tells it not to. A couple of prompts later, if it encounters any error in that same region, it again rips out the typesafe code and replaces it with an `as`.
My observation has been this: if you push a(/any) current-day LLM too close to the edge of its abilities, it goes "insane". Hallucinations start happening everywhere, it stops ignoring previous knowledge, etc. The best way out is to end the session, maybe do some manual work to get to a good state, perhaps update the specs, and start with a fresh context. Using "strong words" or prompting more is of no consequence, the LLM will produce essentially gibberish until reset. Sometimes using a more expensive model temporarily gets around whatever is triggering the stupidity.
I have a product I built that uses some standard automation tools to do order entry into an accounting system. Currently my customer pays people to manually type the orders in from their web portal. The accounting system is closed and they don’t allow easy ways to automate these workflows. Automation is gated behind mega expensive consultants. I’m hoping in the arms race of locking it down to try to prevent 3rd party integration the AI operator model will end up working.
Hard for me to see how it’s ethical to force your customers to do tons of menial data entry when the orders are sitting right there in json.
I routinely flew that far in a plane I spent $100k on. You can land and refuel and a plane that 155+ kts can get quite far in two hops with 3 hour legs. The weather and other maintenance inconveniences is more the issue.
Also, ended up swapping the Pi I started with to a jetson.
reply