> This is false. Launch on Falcon 9 is under $1K/kg.
Every number I find is ~$2,700 or so per kg. I'd be interested to see your data.
>$10/kg is a reasonable expectation.
Citation needed. The tyranny of the rocket equation means you need about 95kg of propellant for every kg of payload. To be clear, assuming it costs $0 to build a rocket, you're going to need to find propellant for under 10c/kg.
Or flip it around. Falcon 9 can lift 18,400kg. You're suggesting therefore that the gross income for spacex will be $184K per launch at $10/kg. It currently costs them about $15-20M per launch. They have to reduce their cost 100x to make a profit with your numbers. From $15,000,000 to ~$150,000. That seems... unlikely.
> Entry from near-orbital speed? Also survivable passively.
...
To be clear, you're suggesting that a rocket in a bad state (engine failure / tumbling) is able to survive a re-entry at 17,500mph?
Also -- your "armored canister". How many kg of armor per kg of payload? 50:50?
That $1000/kg figure is reprocessing something like 10kg of material to get 1kg of enriched material.
Current costs to launch a kg to orbit are something like $2,000-$6,000.
If we're comparing it to you enriched amount, you'd have to launch 10kg, which means you'd have to hit $100/kg launch costs to break even. I'm not convinced that will ever be possible; that's on the order of a first class arline ticket for a 50kg person.
This also doesn't consider what you're getting for launching it into orbit. If the 100 year risk of a casket leaking is (say) x mrads into the environment, you're going to have to consider what the equivalent of vaporizing all of it into the atmosphere in the event of a failed launch is. the fallout (hah) from a failed launch of nuclear waste seems magnitudes more catastrophic than having the same stuff slowly leaking in the middle of a desert cave.
* 1m² = 1x2m? 4m² = 2x4m? etc. I'm confused by this.
* Why 25 m² (5x10 m) = 20.8kW, but 100 m² (10x20 m) = 62.4kW? 4x the size in m², but only ~3x the power? Shouldn't it be 83.2kW, not 62.4kW? It doesn't make sense efficiency would drop....
* You're speccing this as having a 25 year life (with $50/yr maintenance). I'm feeling a bit doubtful that zip-tied tarps under UV and dynamic load are going to last that long, to be honest. The tracking system also seems extremely susceptible to dirt etc.
* This thing looks pretty janky. I'm also not convinced a good 20-year storm wouldn't completely wreck it. Proper wind turbines have the ability to weather-vane into the wind, lock the rotor, etc etc. Again, in it's current incarnation, I'm doubtful it would survive.
* Your "guaranteed Cp of .32" seems... optimistic. Given it looks like you've built actual units, what are your real-world results vs the CFD numbers?
Yep -- I've gotten multiple postcards in the mail and/or product box after ordering something telling me I'd get ~$5-$10 if I left a 5 star review. They thoughtfully included exact screenshots of how to submit the review.
I also went looking for a way to report it. Eventually opened up a chat, didn't really get anywhere.
(I also didn't appreciate getting junk mail for ordering something).
I've been playing around, got it to work, although quality was a bit crappy. Still playing around with the settings that get exposed, but you're welcome to look at :
https://huggingface.co/spaces/jonatei/MoebiusDemo
Note that I'm actively messing with it, so it may break for short periods of time :)
It's also running on the free CPU, so it's like 80 seconds per image...
In VR, it's well known that your peripheral vision is a cue to your brain to infer motion direction.
You can give yourself serious motion sickness in VR if you grab hold of the VR camera and force a roll / pitch in this state.
One of the things some games do is during motion create "tunnel vision" when you move, shrinking down the size of the video feed and putting a black border around it. This significantly reduces this unpleasant effect. I assume the high resolution portion of your eyeball isn't used for motion inference, which makes sense.
I'm thinking this feature is exploiting this same peripheral vision cue.
Most people aren't in a quiet environment when they listen to music these days. Compression helps significantly with this.
What would be neat would be to have a compression metadata 'guide' that would allow a compressor on-device to perform the compression, rather than baked into the audio track.
This would allow the user to tune 'severity' of compression. In a car / fancy headphones, you could sample the ambient noise level and adjust accordingly.
While that's pithy, I think it's also incorrect, because it implies that Oracle / Ellison is controllable by us, in the same way a tool / lawnmower is. That's absolutely not true. It has its own motivations that are best-case neutral to our goals.
It might be better to think of ourselves as individual fish in a school of fishes, and Oracle is the boat with a mile long dragnet. It doesn't care about the individual fish; it's not worth it's time to consider us individually. It's thinking in terms of tonnage.
Of course. LLMS still have huge weaknesses in distinguishing between incoming unsanitized data, and their operating instructions.
It's still malware though. Unlike some backdoor that you could plausibly claim was just a simple memory leak, the instructions for this one are literally written in plain english. Wouldn't be very difficult to show intent to a jury with that one...
One of the all time greats. I think I'll play through it this evening.
"...And what is the 'Self', if not a pattern of data? What is consciousness, if not an illusion of intelligence residing within meat?"
— Prime Function Aki Zeta-5, "The Fallacies of Self-Awareness"
For such low frequency stuff, it feels way safer to just buy a cheap <$500 scope for this kind of work. Using a $50k scope when it's not needed just seems needlessly risky.
Also, float the DUT, not the scope... Sometimes that's not possible, and the temptation is there, but it's really not worth it. Just buy the right gear like a diff probe. You can get one for a few hundred bucks if you don't mind going downmarket.
You can also use two probes and do CH2 - CH1. (Disconnect the GND clips!)
How is it that professional stethoscopes can be that different, and yet this 3D printed one can match a gold-standard one almost exactly?
From what I can tell there's no audio engineering / modelling that's been done here -- It's just some crude openSCAD tubes. And it's not even optimized for 3D printing; a 3D printed tube with a circular cross-section is going to have bridging issues at the top which will result in internal roughness. I have to imagine that results in attenuation. (A better internal shape for a tube is something that looks like "ô". The ^ will print much better)
The type of plastic used and its frequency response, the thickness / stiffness of the silicone tubing, the height / width of the bell... There are so many variables that I think would make significant differences in performance. The fact that they see basically no difference is highly suspect.
This feels like one of those "3D-print everything" fads that was popular a few years ago. Yes, you can make a 3D-printed adjustable wrench, but even the most miserable dollar-tree metal version will beat it in every possible metric.
Likewise, on Alibaba, if you order 200 pieces, I'm seeing metal ones as low as $1.22/pc. I don't believe that this 3D printed one will even be as good as those.
The point of a benchmark is that it allows a relative comparison. The Pelican is one such benchmark.
Feel free to create a "how does it compare to Claude 3.5 Sonnet" benchmark. If people find it useful, it will be run against new LLMs to generate additional points of comparison.
I will also say; it's really easy to just skim past comments. I suspect your ROI time-wise in creating this account to complain will never be recouped compared with just skimming past pelican comment chains.
Every number I find is ~$2,700 or so per kg. I'd be interested to see your data.
>$10/kg is a reasonable expectation.
Citation needed. The tyranny of the rocket equation means you need about 95kg of propellant for every kg of payload. To be clear, assuming it costs $0 to build a rocket, you're going to need to find propellant for under 10c/kg.
Or flip it around. Falcon 9 can lift 18,400kg. You're suggesting therefore that the gross income for spacex will be $184K per launch at $10/kg. It currently costs them about $15-20M per launch. They have to reduce their cost 100x to make a profit with your numbers. From $15,000,000 to ~$150,000. That seems... unlikely.
> Entry from near-orbital speed? Also survivable passively.
... To be clear, you're suggesting that a rocket in a bad state (engine failure / tumbling) is able to survive a re-entry at 17,500mph? Also -- your "armored canister". How many kg of armor per kg of payload? 50:50?