TurtleCracker's recent activity
-
Comment on Something big is happening in ~tech
-
Comment on Something big is happening in ~tech
TurtleCracker Link ParentAI tools do generate productivity gains, but also losses at my work. AI will often give us code that looks right, but isn’t. It will write tests that test nothing. It will generate code with...AI tools do generate productivity gains, but also losses at my work. AI will often give us code that looks right, but isn’t. It will write tests that test nothing. It will generate code with security issues. It will hallucinate features or dependencies that don’t exist and then apologize after you spend 10 minutes trying to figure it out and prove it wrong. It’s also getting more expensive - we run out of “premium credits” all the time.
This is with latest models.
Given all of the above it usually results in net productivity gain, but not a huge one.
-
Comment on Elon Musk says SpaceX will prioritize a city on the moon instead of a colony on Mars in ~space
TurtleCracker Link ParentThe only thing I can think of where the moon is valuable is some sort of atmospheric event that we didn't have forewarning of. Even then, the moon wouldn't be self sufficient so they'd be doomed...There are very, very few potential disasters that could entirely wipe out life on earth but allow life on Mars or the Moon to survive.
The only thing I can think of where the moon is valuable is some sort of atmospheric event that we didn't have forewarning of. Even then, the moon wouldn't be self sufficient so they'd be doomed anyways.
-
Comment on Voyager Technologies CEO says space data center cooling problem still needs to be solved in ~space
TurtleCracker Link ParentI agree that 10 years is unlikely but I don’t know if 300 is accurate. If I look specifically at US naval vessels over the last 100 years they got 2x longer, 3x heavier, went from steam turbines...I agree that 10 years is unlikely but I don’t know if 300 is accurate. If I look specifically at US naval vessels over the last 100 years they got 2x longer, 3x heavier, went from steam turbines and oil boilers to nuclear reactors.
Airplanes barely existed 100 years ago, they are common now. Likewise computers (as we commonly think of them) have existed for less than a hundred years.
I’m more optimistic that we will make significant progress in space during my lifetime.
-
Comment on Voyager Technologies CEO says space data center cooling problem still needs to be solved in ~space
TurtleCracker Link ParentI guess in my head I don’t envision a foundry attached to a data center. I envision heat sinks being moved from a data center to a foundry. Two completely different facilities.I guess in my head I don’t envision a foundry attached to a data center. I envision heat sinks being moved from a data center to a foundry. Two completely different facilities.
-
Comment on Voyager Technologies CEO says space data center cooling problem still needs to be solved in ~space
TurtleCracker Link ParentI agree with most of your points, but I enjoy the discussion so I'm going to continue. Space elevator on the moon to solve this problem? I believe a lot of research has actually been done on that...I agree with most of your points, but I enjoy the discussion so I'm going to continue.
That would require regularly going up and down a gravity well, which would make the whole thing require an incredible amount of energy.
Space elevator on the moon to solve this problem? I believe a lot of research has actually been done on that topic.
If there ever were a need to offload any of it though, it would be a heck of a lot easier to just have that compute on earth.
Isn't there an issue with transmission cutting out between Earth and vehicles during re-entry and/or ascent? I believe SpaceX has had to swap over to satellite based feeds for signal during some of their launches. I guess if you need real time compute sensitive to that level of signal interruption you wouldn't want it happening remotely anyways.
Any thoughts on compute demands increasing in space in the near future?
-
Comment on Voyager Technologies CEO says space data center cooling problem still needs to be solved in ~space
TurtleCracker (edited )Link ParentIf you have modular heat sinks then you don't need the data center on the moon, you could shift heat to the moon. Would a data center in space allow you to have less compute on the vehicles...You could transfer it to the moon, but that would require the data center to be on the moon, not in space, and the moon is way harder to get to than orbit.
If you have modular heat sinks then you don't need the data center on the moon, you could shift heat to the moon.
There's no economic reason to do it though. Every aspect of it would be more expensive than putting them on earth, and not a little more expensive, like... tens of thousands of times more expensive, not to mention risky and dangerous.
Would a data center in space allow you to have less compute on the vehicles themselves by offloading the work?
-
Comment on Voyager Technologies CEO says space data center cooling problem still needs to be solved in ~space
TurtleCracker Link ParentDoes it have to be radiated outward, or is that what we currently do?* It seems like if you can get it into contact with something colder than the heat, it could be used for a beneficial process....Unfortunately, that transfer is the very thing that's difficult in space. Since there's no direct contact with the environment to cool via conduction or convection, heat has to be radiated outwards – essentially by ensuring that the device's infrared glow points away from the sun. That's not very efficient, particularly when the solar panels have to be pointed at the sun to provide power.
Does it have to be radiated outward, or is that what we currently do?* It seems like if you can get it into contact with something colder than the heat, it could be used for a beneficial process. If we bring ore from asteroids into a facility - couldn't we use heat to process that material into something more useful internal to the facility itself? Couldn't we channel the heat into a sink and then transfer that to something else - maybe the moon? Transfer the heat into the moon?
I'm not an expert. Just poking at it. We have plenty of orbital bodies that don't seem to have issues managing heat.
*I understand that ultimately it all results in radiating outward, no matter how many steps in between.
-
Comment on Voyager Technologies CEO says space data center cooling problem still needs to be solved in ~space
TurtleCracker Link ParentI'm pretty sure we can make modular heat sinks that could be swapped. I agree it's a bit of a chicken/egg problem. We've run into this on earth as well with things like EV infrastructure and other...I'm pretty sure we can make modular heat sinks that could be swapped. I agree it's a bit of a chicken/egg problem. We've run into this on earth as well with things like EV infrastructure and other technologies that require some sort of secondary thing to exist to enable them to be more efficient / cheaper.
-
Comment on Voyager Technologies CEO says space data center cooling problem still needs to be solved in ~space
TurtleCracker Link ParentAre the launch costs as relevant for infrastructure already in space? If I launch some sort of base specifically designed to collect and process/radiate waste heat from other facilities for a fee...Are the launch costs as relevant for infrastructure already in space? If I launch some sort of base specifically designed to collect and process/radiate waste heat from other facilities for a fee for example.
-
Comment on Voyager Technologies CEO says space data center cooling problem still needs to be solved in ~space
TurtleCracker Link ParentI was talking about waste heat as well. I'm not an expert, just interested in the topic. Heat is energy. If heat is plentiful and cheap in space, it seems like an economic opportunity to find a...I was talking about waste heat as well. I'm not an expert, just interested in the topic. Heat is energy. If heat is plentiful and cheap in space, it seems like an economic opportunity to find a way to turn that energy into something valuable. This happens a lot with manufacturing processes where we turn something that used to be wasted into something valuable.
-
Comment on Voyager Technologies CEO says space data center cooling problem still needs to be solved in ~space
TurtleCracker (edited )Link ParentPreamble: I am not an expert in any of this. Just interested in it. You can use anisotropic radiation of waste heat to provide a tiny amount of thrust. See: Support for the thermal origin of the...Preamble: I am not an expert in any of this. Just interested in it.
No. Thrust requires a propellant, not heat.
You can use anisotropic radiation of waste heat to provide a tiny amount of thrust. See: Support for the thermal origin of the Pioneer anomaly.
I was also able to find EQUULEUS which appears to use the waste heat from communications equipment in part of it's propulsion systems.
No, generating energy from heat requires both a hot and a comparatively very cold region, so if anything you would need more cooling, plus an entire turbine apparatus, plus "using" the heat to generate power wouldn't actually make it go anywhere, it would still be hot as hell
Doing some basic research on this I seemed to hit three topics:
Radioisotope thermoelectric generator (RTG)
Stirling Engines
Brayton CyclesI'm not sure if these are explicitly using waste heat, but many of them consume the heat generated by some other power generation process and recover that to generate more electricity.
Though you are just moving the forms of energy around with this. IE if you charge a battery, it will eventually release most of that energy back out as heat anyways. It's just more useful to have electricity instead of heat.
I don't understand how heat would be relevant to manufacturing besides the energy angle I've already addressed
Smelting? Welding? Refining? This is the topic I have done the least research into. We use heat to do these things on earth. If we mine other planets and/or asteroids it may make more sense for a lot of the refining and manufacturing of those raw materials to happen in space before moving back into the atmosphere. Especially if waste heat is a cheap and plentiful source of energy.
Can you elaborate on what you mean by this? I can think of some ways to store heat but none of them would actually make a spacebound facility colder.
Pump the heat into a sink or storage device (molten salt? etc?). Another vessel swaps the sink with a cold sink and transports the heat to somewhere it can be used for something beneficial.
-
Comment on Voyager Technologies CEO says space data center cooling problem still needs to be solved in ~space
TurtleCracker LinkCan’t the heat be used for thrust, power generation, manufacturing, or life support? The problem seems to be due to limited infrastructure that demands heat in space. Heat is energy. It’s easy to...Can’t the heat be used for thrust, power generation, manufacturing, or life support? The problem seems to be due to limited infrastructure that demands heat in space. Heat is energy. It’s easy to store heat in space.
-
Comment on Building a C compiler with a team of parallel Claudes in ~tech
TurtleCracker Link ParentMaybe? But that’s not what clean-room means. The equivalent for humans is if you let engineers read compiler source code for a few months before locking them in a room with no internet and had...Maybe? But that’s not what clean-room means. The equivalent for humans is if you let engineers read compiler source code for a few months before locking them in a room with no internet and had them make a compiler. It’s as close as you can get to an open book test without being one.
-
Comment on Building a C compiler with a team of parallel Claudes in ~tech
TurtleCracker LinkIsn't this disingenuous? Presumably the model was trained on existing compilers and codebases. How can it be considered clean-room?This was a clean-room implementation (Claude did not have internet access at any point during its development); it depends only on the Rust standard library.
Isn't this disingenuous? Presumably the model was trained on existing compilers and codebases. How can it be considered clean-room?
-
Comment on Any software engineers considering a career switch due to AI? in ~comp
TurtleCracker Link ParentI'm not talking exclusively about journalists. I'm also talking about investors and boards. A pretty large amount of time is often spent gathering data, framing data, and doing presentations to...I'm not talking exclusively about journalists. I'm also talking about investors and boards. A pretty large amount of time is often spent gathering data, framing data, and doing presentations to boards and/or investors. Depends on the company structure and size. I get your point though.
-
Comment on Any software engineers considering a career switch due to AI? in ~comp
TurtleCracker Link ParentI've experienced LLMs frequently writing unit tests where the mocking/stubbing they do in setting up the test effectively means they are only testing the mocking and stubbing. The test passes, it...I've experienced LLMs frequently writing unit tests where the mocking/stubbing they do in setting up the test effectively means they are only testing the mocking and stubbing. The test passes, it looks great, but it's not actually testing anything.
I've also seen some pretty significant security issues show up in code review processes because of AI generated code.
-
Comment on Any software engineers considering a career switch due to AI? in ~comp
TurtleCracker Link ParentIf you can announce publicly "We are laying off 1,000 people because we can pay someone in another country 10% of their wage." versus "We are laying off 1,000 people because of our innovative use...If you can announce publicly "We are laying off 1,000 people because we can pay someone in another country 10% of their wage." versus "We are laying off 1,000 people because of our innovative use of artificial intelligence." which do you think they'd choose for framing the message?
-
Comment on Any software engineers considering a career switch due to AI? in ~comp
TurtleCracker LinkI view LLMs more as tools that accelerate software engineering, not replace it. Like compilers, linting, static analysis, APMs, or IDEs. Business and sales people still need someone to use the...I view LLMs more as tools that accelerate software engineering, not replace it. Like compilers, linting, static analysis, APMs, or IDEs.
Business and sales people still need someone to use the tools. Interpret their desires in a sensible way. Avoid pitfalls. Debug issues, etc.. A product owner or technical project manager or something might be able to do a lot of this, but will still probably fall short in areas.
I don’t know a single engineer that has lost their job to be replaced by AI. I have heard of engineers that lost their jobs because of the budget companies are spending on AI - but that’s different from being replaced by it.
I also suspect many companies are using AI as a convenient excuse for layoffs that are actually driven by the economy, politics, or offshoring.
-
Comment on Alphabet plots big expansion in India as US restricts visas in ~tech
TurtleCracker Link ParentI think a team in India can make a lot of sense if you are doing it for the right reasons with the right support structure. That doesn’t seem to be what happens though. You should go to India for...I think a team in India can make a lot of sense if you are doing it for the right reasons with the right support structure. That doesn’t seem to be what happens though. You should go to India for the time zone coverage and talent, not strictly cost savings.
Teams in India have many hidden costs. If you are offshoring to save money only, you probably won’t save much money. You’ll end up spending more on communication, vision/product alignment, and rework.
This kind of thing is really frustrating as a default behavior. The output is always way too verbose. You have to practically berate some of the models to get them to tighten up the output.