Elon Musk quote: SpaceX will lean in big on the Moon

[OTV Booster]

But I've learned over he years that when Musk says stuff, even if counter-intuitive to me  I should listen.

He's not right 100% of the time, but he's up there above 90, which is not bad for a visionary.

However, I think AI type things are one of his specific weak points. I still remember 'we will have total self driving in 3 years, the regulations just need to catch up' in 2015. In my experience, Musk is historically overoptimistic about the near term potential of AI/robotics/etc.

This is significantly different from rocket and satellite stuff, which is far more predictable. With the exception of a few questions like supersonic retropropulsion, it was clear long ago (since at least DC-X/Masten) that a system like F9 could be built. The question was instead whether the market demand would be there to justify it.

Similar for Starlink -  I don't think anyone questioned the technical possibility of such a system, only its finances.

It is in contrast far from clear whether there is any role for LLM style AI (as distinctly opposed to specialized systems like AlphaFold) which justify trillion dollar valuations. Google has gotten IMO clearly worse as they have integrated LLMs.

this whole idea is based on the premise that "thinking interfaces" will be everywhere.  Every happy door, every smart elevator, every empath coffee maker.

Why would you want that though? What's the advantage?

Specialized tools (like a door or a coffee maker) don't need a generalized LLM type AI. A door just needs to open/close, a coffee maker just needs to make coffee. Even remote operation (turn on the coffee warmer from your phone) doesn't need that

Personally?  I'd love conversational AI-based interfaces instead of rigid algorithmic ones that can't deal with any input that's not precisely structured

IDK. For computers / smartphones maybe*. But for household objects like doors and coffee makers? I see zero advantage.

I feel like this is kind of a lose lose situation. If it IS a bubble, the economy is messed up. If it's NOT a bubble, a lot of jobs get lost with no clear replacement, and people get more and more used to getting quick easy answers from very "black box" tech whose hidden biases & filters are not at all visible to the user and can be changed by the company controlling it.

I don't see what the benefit outweighing all this (to people in general, not to the company's valuation) is supposed to be.

*And even then, there's a *huge* risk in "too human appearing" conversational tech messing up younger or psychologically vulnerable people. I personally tend to think that this issue makes conversational AI a net negative in terms of human well-being, even if it's a money maker.

But this is probably well off topic. My point was about a possible bubble burst destroying the funds for Mars, not really about the value of AI in general. (A technology can be very useful and *Still* have a bubble early on; the Internet did.)

I want to get into my car and have an intelligent conversation with it about where I'm going, and have it co-pilot me.  Not the BS maps interface I get today.

Same with my house.  This is not "Siri turn on the lights" kind of stuff.

And chatGPT today is 10x more intelligent than the customer care rep I get on the line at my bank or airline.

And I haven't even touched the intersection of AI and robotics.

There's a whole universe of what AI can do, I'm sure I'm not imagining even a percent of it.

Man, I am sorry but I have to dis you. You want an intelligent conversation with your car? I've come to respect you too much to accept this. Good prompts on where to turn, but conversation...?


As for the phone people at the bank... I needed to find out the source of a cashiers check received by me that originated from a credit union of which I am not a member. All phone numbers let to a commercial call center that could only answer routine questions like "what is my balance?" The best of AI can't compensate for poor business practices.

[meekGee]

But I've learned over he years that when Musk says stuff, even if counter-intuitive to me  I should listen.

He's not right 100% of the time, but he's up there above 90, which is not bad for a visionary.

However, I think AI type things are one of his specific weak points. I still remember 'we will have total self driving in 3 years, the regulations just need to catch up' in 2015. In my experience, Musk is historically overoptimistic about the near term potential of AI/robotics/etc.

This is significantly different from rocket and satellite stuff, which is far more predictable. With the exception of a few questions like supersonic retropropulsion, it was clear long ago (since at least DC-X/Masten) that a system like F9 could be built. The question was instead whether the market demand would be there to justify it.

Similar for Starlink -  I don't think anyone questioned the technical possibility of such a system, only its finances.

It is in contrast far from clear whether there is any role for LLM style AI (as distinctly opposed to specialized systems like AlphaFold) which justify trillion dollar valuations. Google has gotten IMO clearly worse as they have integrated LLMs.

this whole idea is based on the premise that "thinking interfaces" will be everywhere.  Every happy door, every smart elevator, every empath coffee maker.

Why would you want that though? What's the advantage?

Specialized tools (like a door or a coffee maker) don't need a generalized LLM type AI. A door just needs to open/close, a coffee maker just needs to make coffee. Even remote operation (turn on the coffee warmer from your phone) doesn't need that

Personally?  I'd love conversational AI-based interfaces instead of rigid algorithmic ones that can't deal with any input that's not precisely structured

IDK. For computers / smartphones maybe*. But for household objects like doors and coffee makers? I see zero advantage.

I feel like this is kind of a lose lose situation. If it IS a bubble, the economy is messed up. If it's NOT a bubble, a lot of jobs get lost with no clear replacement, and people get more and more used to getting quick easy answers from very "black box" tech whose hidden biases & filters are not at all visible to the user and can be changed by the company controlling it.

I don't see what the benefit outweighing all this (to people in general, not to the company's valuation) is supposed to be.

*And even then, there's a *huge* risk in "too human appearing" conversational tech messing up younger or psychologically vulnerable people. I personally tend to think that this issue makes conversational AI a net negative in terms of human well-being, even if it's a money maker.

But this is probably well off topic. My point was about a possible bubble burst destroying the funds for Mars, not really about the value of AI in general. (A technology can be very useful and *Still* have a bubble early on; the Internet did.)

I want to get into my car and have an intelligent conversation with it about where I'm going, and have it co-pilot me.  Not the BS maps interface I get today.

Same with my house.  This is not "Siri turn on the lights" kind of stuff.

And chatGPT today is 10x more intelligent than the customer care rep I get on the line at my bank or airline.

And I haven't even touched the intersection of AI and robotics.

There's a whole universe of what AI can do, I'm sure I'm not imagining even a percent of it.

Man, I am sorry but I have to dis you. You want an intelligent conversation with your car? I've come to respect you too much to accept this. Good prompts on where to turn, but conversation...?


As for the phone people at the bank... I needed to find out the source of a cashiers check received by me that originated from a credit union of which I am not a member. All phone numbers let to a commercial call center that could only answer routine questions like "what is my balance?" The best of AI can't compensate for poor business practices.

Of course.

Most of the frustration with automation is just how dumb it is.

Most of the attraction with using chat while engineering is that there's context, and measure, and calibration.  There's even a level of independence in that things come up that I didn't directly ask for.

I don't want to phrase queries exactly the way Google maps expects them, and then have to filter out 90% of garbage results including shops that are behind me, or will be closed, etc. I want to have a natural language conversation about it.

Or say that I'm out of XXX and YYY and have the car figure out how to best get it.  But if I'm on the East coast, it won't try to send me to a Home Depot in Utah because it's on the way.

Basically I want to talk to it like it was a cognizant human being. With wheels and a motor.

For example, it sees all around it, right?  So if an interesting truck drives by, I want to be able to ask WTF was that, and then be able to pull on that string.

I want to be able to comment on an odd looking cloud, or a plane that caught my eye...

Basically the kinda conversation I'd have with a person, except of course colored by the limitations and potential of an AI.

Yup my plastic pal who's fun to be with.

[OTV Booster]

I thought there was an NSF thread or conversation that was looking at mass drivers on the Moon, and they determined (or came to the conclusion) that in reality they really don't work as well as we thought.

For instance, whatever mass you launch is either going to be semi-orbital (which means you have to have a catcher in orbit that can snatch the payload before it falls back to the Moon) or it leaves the orbit of the Moon, in which case you have to go chase it somehow, but it is difficult to have a "catcher" prepositioned (something to do with unstable Moon orbits or??).

In any case, I think the burden of proof is on those that think we can "manufacture" anything on the Moon, much less something as high tech as solar cells in large volume (as well as everything it takes to install those panels so they can be useful), at any point in near future. I think it is more likely we could have a small Earth-supplied colony on Mars well before we could figure out solar cell manufacturing on the Moon.

Blue have already demostrated they can manufacture a solar cell from regolith (artificial). Extracted silicon by electrolysis and then produced cell from it.

Sure, Blue Origin says:

Using regolith simulants, our reactor produces iron, silicon, and aluminum through molten regolith electrolysis, in which an electrical current separates those elements from the oxygen to which they are bound. Oxygen for propulsion and life support is a byproduct.

I've never doubted the materials needed were on the Moon, my whole point is that refining the needed material from raw lunar material, and then transforming that material into finished products, takes a factory.

Not like some box that you can carry, but here on Earth a solar cell manufacturing facility for high volume production would be the size of an American football field. So not only do you have to ship all that high tech equipment to the Moon, then install it and calibrate it, but how do you maintain such delicate machinery? Ooh, and the chicken and egg challenge is how do you power such a factory if it can't yet product the solar cells needed to power it? 

And no, if we can't fully automate a factory here on Earth, then we won't be able to do it on the Moon. Which is why I think a lot of people are not understanding how hard it will be to build solar cells on the Moon.

Spinlaunch has already produced operational mass driver capable of 5000mph. Lunar escape velocity is 5325mph.

If you look at what I wrote, I didn't dispute that mass could be thrown into space. My post talked about the challenge of CATCHING whatever was launched. How do you stop something that is traveling X kph? What happens to the kinetic energy? How exact does the launch need to be in order for the payload to be caught? THAT is what I remember NSF members talking about, and how they thought that would be extremely difficult.

Assume launch is from a lunar pole. Launch from a mass driver with apolune perilune at or near EML1. Relatively small onboard thrusters or a local tug would be enough to put it in a halo orbit.

Alternatively, launch with a sling throwing a counterweight in the opposite direction. The counterweight would need quite a bit more dV to work its way around to L1. Maybe not worth it. Maybe some creative choices of mass and arm length could get it to L2? At minimum this would keep it from impacting back on the moon.

Neither will happen in the near future.


Edit: changed word

[OTV Booster]

But I've learned over he years that when Musk says stuff, even if counter-intuitive to me  I should listen.

He's not right 100% of the time, but he's up there above 90, which is not bad for a visionary.

However, I think AI type things are one of his specific weak points. I still remember 'we will have total self driving in 3 years, the regulations just need to catch up' in 2015. In my experience, Musk is historically overoptimistic about the near term potential of AI/robotics/etc.

This is significantly different from rocket and satellite stuff, which is far more predictable. With the exception of a few questions like supersonic retropropulsion, it was clear long ago (since at least DC-X/Masten) that a system like F9 could be built. The question was instead whether the market demand would be there to justify it.

Similar for Starlink -  I don't think anyone questioned the technical possibility of such a system, only its finances.

It is in contrast far from clear whether there is any role for LLM style AI (as distinctly opposed to specialized systems like AlphaFold) which justify trillion dollar valuations. Google has gotten IMO clearly worse as they have integrated LLMs.

this whole idea is based on the premise that "thinking interfaces" will be everywhere.  Every happy door, every smart elevator, every empath coffee maker.

Why would you want that though? What's the advantage?

Specialized tools (like a door or a coffee maker) don't need a generalized LLM type AI. A door just needs to open/close, a coffee maker just needs to make coffee. Even remote operation (turn on the coffee warmer from your phone) doesn't need that

Personally?  I'd love conversational AI-based interfaces instead of rigid algorithmic ones that can't deal with any input that's not precisely structured

IDK. For computers / smartphones maybe*. But for household objects like doors and coffee makers? I see zero advantage.

I feel like this is kind of a lose lose situation. If it IS a bubble, the economy is messed up. If it's NOT a bubble, a lot of jobs get lost with no clear replacement, and people get more and more used to getting quick easy answers from very "black box" tech whose hidden biases & filters are not at all visible to the user and can be changed by the company controlling it.

I don't see what the benefit outweighing all this (to people in general, not to the company's valuation) is supposed to be.

*And even then, there's a *huge* risk in "too human appearing" conversational tech messing up younger or psychologically vulnerable people. I personally tend to think that this issue makes conversational AI a net negative in terms of human well-being, even if it's a money maker.

But this is probably well off topic. My point was about a possible bubble burst destroying the funds for Mars, not really about the value of AI in general. (A technology can be very useful and *Still* have a bubble early on; the Internet did.)

I want to get into my car and have an intelligent conversation with it about where I'm going, and have it co-pilot me.  Not the BS maps interface I get today.

Same with my house.  This is not "Siri turn on the lights" kind of stuff.

And chatGPT today is 10x more intelligent than the customer care rep I get on the line at my bank or airline.

And I haven't even touched the intersection of AI and robotics.

There's a whole universe of what AI can do, I'm sure I'm not imagining even a percent of it.

Man, I am sorry but I have to dis you. You want an intelligent conversation with your car? I've come to respect you too much to accept this. Good prompts on where to turn, but conversation...?


As for the phone people at the bank... I needed to find out the source of a cashiers check received by me that originated from a credit union of which I am not a member. All phone numbers let to a commercial call center that could only answer routine questions like "what is my balance?" The best of AI can't compensate for poor business practices.

Of course.

Most of the frustration with automation is just how dumb it is.

Most of the attraction with using chat while engineering is that there's context, and measure, and calibration.  There's even a level of independence in that things come up that I didn't directly ask for.

I don't want to phrase queries exactly the way Google maps expects them, and then have to filter out 90% of garbage results including shops that are behind me, or will be closed, etc. I want to have a natural language conversation about it.

Or say that I'm out of XXX and YYY and have the car figure out how to best get it.  But if I'm on the East coast, it won't try to send me to a Home Depot in Utah because it's on the way.

Basically I want to talk to it like it was a cognizant human being. With wheels and a motor.

For example, it sees all around it, right?  So if an interesting truck drives by, I want to be able to ask WTF was that, and then be able to pull on that string.

I want to be able to comment on an odd looking cloud, or a plane that caught my eye...

Basically the kinda conversation I'd have with a person, except of course colored by the limitations and potential of an AI.

Yup my plastic pal who's fun to be with.

OMG. You want a relationship with your car? I think there might be a generational disjuncture here. I just want a car to run straight, stop straight and keep the water out.

[Eka]

IDK. For computers / smartphones maybe*. But for household objects like doors and coffee makers? I see zero advantage.

I feel like this is kind of a lose lose situation. If it IS a bubble, the economy is messed up. If it's NOT a bubble, a lot of jobs get lost with no clear replacement, and people get more and more used to getting quick easy answers from very "black box" tech whose hidden biases & filters are not at all visible to the user and can be changed by the company controlling it.

I don't see what the benefit outweighing all this (to people in general, not to the company's valuation) is supposed to be.

*And even then, there's a *huge* risk in "too human appearing" conversational tech messing up younger or psychologically vulnerable people. I personally tend to think that this issue makes conversational AI a net negative in terms of human well-being, even if it's a money maker.

But this is probably well off topic. My point was about a possible bubble burst destroying the funds for Mars, not really about the value of AI in general. (A technology can be very useful and *Still* have a bubble early on; the Internet did.)

I want to get into my car and have an intelligent conversation with it about where I'm going, and have it co-pilot me.  Not the BS maps interface I get today.

Same with my house.  This is not "Siri turn on the lights" kind of stuff.

And chatGPT today is 10x more intelligent than the customer care rep I get on the line at my bank or airline.

And I haven't even touched the intersection of AI and robotics.

There's a whole universe of what AI can do, I'm sure I'm not imagining even a percent of it.

Man, I am sorry but I have to dis you. You want an intelligent conversation with your car? I've come to respect you too much to accept this. Good prompts on where to turn, but conversation...?


As for the phone people at the bank... I needed to find out the source of a cashiers check received by me that originated from a credit union of which I am not a member. All phone numbers let to a commercial call center that could only answer routine questions like "what is my balance?" The best of AI can't compensate for poor business practices.

Of course.

Most of the frustration with automation is just how dumb it is.

Most of the attraction with using chat while engineering is that there's context, and measure, and calibration.  There's even a level of independence in that things come up that I didn't directly ask for.

I don't want to phrase queries exactly the way Google maps expects them, and then have to filter out 90% of garbage results including shops that are behind me, or will be closed, etc. I want to have a natural language conversation about it.

Or say that I'm out of XXX and YYY and have the car figure out how to best get it.  But if I'm on the East coast, it won't try to send me to a Home Depot in Utah because it's on the way.

Basically I want to talk to it like it was a cognizant human being. With wheels and a motor.

For example, it sees all around it, right?  So if an interesting truck drives by, I want to be able to ask WTF was that, and then be able to pull on that string.

I want to be able to comment on an odd looking cloud, or a plane that caught my eye...

Basically the kinda conversation I'd have with a person, except of course colored by the limitations and potential of an AI.

Yup my plastic pal who's fun to be with.

OMG. You want a relationship with your car? I think there might be a generational disjuncture here. I just want a car to run straight, stop straight and keep the water out.

Yeah, there is a good reason the Marketing Department of the Sirius Cybernetics Corporation were the first against the wall when the revolution came.

I just want my car to drive me wherever I want to go so I can sleep as it does it.

I don't think the sane ones of us will be able to stop it. AIs will get shoved into everything. At some point in the future it will be near impossible to get a simple microcontroller without an AI built into it.

[Eka]

I thought there was an NSF thread or conversation that was looking at mass drivers on the Moon, and they determined (or came to the conclusion) that in reality they really don't work as well as we thought.

For instance, whatever mass you launch is either going to be semi-orbital (which means you have to have a catcher in orbit that can snatch the payload before it falls back to the Moon) or it leaves the orbit of the Moon, in which case you have to go chase it somehow, but it is difficult to have a "catcher" prepositioned (something to do with unstable Moon orbits or??).

In any case, I think the burden of proof is on those that think we can "manufacture" anything on the Moon, much less something as high tech as solar cells in large volume (as well as everything it takes to install those panels so they can be useful), at any point in near future. I think it is more likely we could have a small Earth-supplied colony on Mars well before we could figure out solar cell manufacturing on the Moon.

Blue have already demostrated they can manufacture a solar cell from regolith (artificial). Extracted silicon by electrolysis and then produced cell from it.

Sure, Blue Origin says:

Using regolith simulants, our reactor produces iron, silicon, and aluminum through molten regolith electrolysis, in which an electrical current separates those elements from the oxygen to which they are bound. Oxygen for propulsion and life support is a byproduct.

I've never doubted the materials needed were on the Moon, my whole point is that refining the needed material from raw lunar material, and then transforming that material into finished products, takes a factory.

Not like some box that you can carry, but here on Earth a solar cell manufacturing facility for high volume production would be the size of an American football field. So not only do you have to ship all that high tech equipment to the Moon, then install it and calibrate it, but how do you maintain such delicate machinery? Ooh, and the chicken and egg challenge is how do you power such a factory if it can't yet product the solar cells needed to power it? 

And no, if we can't fully automate a factory here on Earth, then we won't be able to do it on the Moon. Which is why I think a lot of people are not understanding how hard it will be to build solar cells on the Moon.

Spinlaunch has already produced operational mass driver capable of 5000mph. Lunar escape velocity is 5325mph.

If you look at what I wrote, I didn't dispute that mass could be thrown into space. My post talked about the challenge of CATCHING whatever was launched. How do you stop something that is traveling X kph? What happens to the kinetic energy? How exact does the launch need to be in order for the payload to be caught? THAT is what I remember NSF members talking about, and how they thought that would be extremely difficult.

Assume launch is from a lunar pole. Launch from a mass driver with apolune perilune at or near EML1. Relatively small onboard thrusters or a local tug would be enough to put it in a halo orbit.

Alternatively, launch with a sling throwing a counterweight in the opposite direction. The counterweight would need quite a bit more dV to work its way around to L1. Maybe not worth it. Maybe some creative choices of mass and arm length could get it to L2? At minimum this would keep it from impacting back on the moon.

Neither will happen in the near future.


Edit: changed word

I can see how to do it. Launch to orbit and use thusters to put into a proper orbit. Kinda like SpaceX does with Starlinks. Plenty of trace elements will come out of the regolith to provide propellants.

Factories: Bootstrap. Smaller factories build the raw materials and components needed to assemble larger factories. Lather, rinse, repeat. Eventually massive levels of production will be reached with few human overseers. We didn't build our current industrial base overnight. It's been an exponential curve of increasing sophistication and capacity. We will be able to go through it much faster on the Moon and Mars.

[meekGee]

Man, I am sorry but I have to dis you. You want an intelligent conversation with your car? I've come to respect you too much to accept this. Good prompts on where to turn, but conversation...?

As for the phone people at the bank... I needed to find out the source of a cashiers check received by me that originated from a credit union of which I am not a member. All phone numbers let to a commercial call center that could only answer routine questions like "what is my balance?" The best of AI can't compensate for poor business practices.

Of course.

Most of the frustration with automation is just how dumb it is.

Most of the attraction with using chat while engineering is that there's context, and measure, and calibration.  There's even a level of independence in that things come up that I didn't directly ask for.

I don't want to phrase queries exactly the way Google maps expects them, and then have to filter out 90% of garbage results including shops that are behind me, or will be closed, etc. I want to have a natural language conversation about it.

Or say that I'm out of XXX and YYY and have the car figure out how to best get it.  But if I'm on the East coast, it won't try to send me to a Home Depot in Utah because it's on the way.

Basically I want to talk to it like it was a cognizant human being. With wheels and a motor.

For example, it sees all around it, right?  So if an interesting truck drives by, I want to be able to ask WTF was that, and then be able to pull on that string.

I want to be able to comment on an odd looking cloud, or a plane that caught my eye...

Basically the kinda conversation I'd have with a person, except of course colored by the limitations and potential of an AI.

Yup my plastic pal who's fun to be with.

OMG. You want a relationship with your car? I think there might be a generational disjuncture here. I just want a car to run straight, stop straight and keep the water out.

Yeah, old body, young spirit, the secret to my success... 

In seriousness, it's not a relationship in the sense of a human one, but it's interacting with a computer in a human like way, where I don't have to slow myself down and adjust to its dumbness.

And if a chat with your car makes you uncomfortable, think of it as your personal copilot that happens to inhabit the car when you're in it, but goes with you when you get out.

So minus the copilot, we're back to run smooth, turn aharp, and keep the water on the right side of the windshield...

Easier?


(BTW your example with the call center - those piss poor systems are horrible, but are also miles from what good AI can do, even today. Personal chatGPT isn't like that at all.)

[InterestedEngineer]

chatbots is all good, but agents are for some tasks an order of magnitude improvement in productivity over chat bots.

Thread here (where it really belongs):  https://forum.nasaspaceflight.com/index.php?topic=57207.msg2745416#msg2745416


I'm not sure this applies to say a car, but it will apply to many tasks.

Right now some of the tasks described in that thread take 2-3 minutes for the AI to complete (hey it's generating an entire daemon that coordinates two databases with two APIs A human would take hours)

Now, in a desperate attempt to to get this on topic of the thread... oh wait,, this is the moon thread?  I thought it was the "other" AI thread, so now we have three AI threads (at least).


Anyway...  getting that down to 2-3 seconds would be another win in productivity, and that will require space based AI because Moore's Law is no longer delivering very muich and it's just how much power can you apply and cool.

I apologize I can't get the topic of "the moon" into this in any reasonable fashion.  At least I got it space related 

[meekGee]

chatbots is all good, but agents are for some tasks an order of magnitude improvement in productivity over chat bots.

Thread here (where it really belongs):  https://forum.nasaspaceflight.com/index.php?topic=57207.msg2745416#msg2745416


I'm not sure this applies to say a car, but it will apply to many tasks.

Right now some of the tasks described in that thread take 2-3 minutes for the AI to complete (hey it's generating an entire daemon that coordinates two databases with two APIs A human would take hours)

Now, in a desperate attempt to to get this on topic of the thread... oh wait,, this is the moon thread?  I thought it was the "other" AI thread, so now we have three AI threads (at least).


Anyway...  getting that down to 2-3 seconds would be another win in productivity, and that will require space based AI because Moore's Law is no longer delivering very muich and it's just how much power can you apply and cool.

I apologize I can't get the topic of "the moon" into this in any reasonable fashion.  At least I got it space related 

It's moon related because moon depends on a huge increase in power consumption, so applications matter.

Today's chat already does a number of things.

It talks naturally, and keeps context in complicated conversations so you can say "it" and "they" and it figures it out even of the context is confusing.

It reasons.

It researches external sources that it doesn't have "memorized".

However, it is still a prompt-response model.  It doesn't stay "on" between exchanges.

But for example the "follow video" mode does do that, constantly digesting a video feed and commenting when it feels like it.  (But then falls into the ping pong mode for the actual conversation)

How long before they have one that's always on, and can actually manage real time conversing, pauses, interruptions, etc?

Is that chat 6.0 or 7.0?   I think it's an obvious next step  not far fetched, and probably an energy hog.

But once you have a billion of those, let's say consuming a kWatt, you can see where it's going.

(And yes, how to handle latency, I know)

[steveleach]

It reasons.

It researches external sources that it doesn't have "memorized".

Drifting way off topic here, but careful with these two, specifically.

It would be more accurate to say that it statistically predicts what a person who reasons would say, and it presents answers in a format that looks like a person has done research.

But this is a spaceflight forum not an AI forum, so I should probably talk about SpaceX's plans for the moon instead.

[meekGee]

It reasons.

It researches external sources that it doesn't have "memorized".

Drifting way off topic here, but careful with these two, specifically.

It would be more accurate to say that it statistically predicts what a person who reasons would say, and it presents answers in a format that looks like a person has done research.

But this is a spaceflight forum not an AI forum, so I should probably talk about SpaceX's plans for the moon instead.

It's still an interesting topic amd indirectly relevant.

It effectively reasons. It answers "what if" questions rather well, keeping track of what influences what. Infrequently, it gets things wrong.  So do people.

It can look at a single picture, and unprompted, can pick up what's interesting and how it'll likely develop, with what consequences.

Of you're not sure, when it's done answering, just ask him "why" or "how come". Or push back.

I'm sure it'll get better, but it already reasons.

As for research, when sources are available, it woll dive deeper as a result of what it already knows. Sometimes it actually shows the progress as it reads up standards or similar documents.  It's not a smoke show.

[OTV Booster]

Man, I am sorry but I have to dis you. You want an intelligent conversation with your car? I've come to respect you too much to accept this. Good prompts on where to turn, but conversation...?

As for the phone people at the bank... I needed to find out the source of a cashiers check received by me that originated from a credit union of which I am not a member. All phone numbers let to a commercial call center that could only answer routine questions like "what is my balance?" The best of AI can't compensate for poor business practices.

Of course.

Most of the frustration with automation is just how dumb it is.

Most of the attraction with using chat while engineering is that there's context, and measure, and calibration.  There's even a level of independence in that things come up that I didn't directly ask for.

I don't want to phrase queries exactly the way Google maps expects them, and then have to filter out 90% of garbage results including shops that are behind me, or will be closed, etc. I want to have a natural language conversation about it.

Or say that I'm out of XXX and YYY and have the car figure out how to best get it.  But if I'm on the East coast, it won't try to send me to a Home Depot in Utah because it's on the way.

Basically I want to talk to it like it was a cognizant human being. With wheels and a motor.

For example, it sees all around it, right?  So if an interesting truck drives by, I want to be able to ask WTF was that, and then be able to pull on that string.

I want to be able to comment on an odd looking cloud, or a plane that caught my eye...

Basically the kinda conversation I'd have with a person, except of course colored by the limitations and potential of an AI.

Yup my plastic pal who's fun to be with.

OMG. You want a relationship with your car? I think there might be a generational disjuncture here. I just want a car to run straight, stop straight and keep the water out.

Yeah, old body, young spirit, the secret to my success... 

In seriousness, it's not a relationship in the sense of a human one, but it's interacting with a computer in a human like way, where I don't have to slow myself down and adjust to its dumbness.

And if a chat with your car makes you uncomfortable, think of it as your personal copilot that happens to inhabit the car when you're in it, but goes with you when you get out.

So minus the copilot, we're back to run smooth, turn aharp, and keep the water on the right side of the windshield...

Easier?


(BTW your example with the call center - those piss poor systems are horrible, but are also miles from what good AI can do, even today. Personal chatGPT isn't like that at all.)

We are in such totally different wavelengths. First thing I do with a computer is shut off the sound, yank out as much cruft as I can, disable most everything in startup then only plug into the internet as needed.  If my car started talking to me I'd pull over at a pawn shop, buy a gun and shoot it.


I'm in the market for an SUV. First entry in the decision tree is: Does it have a modem? If yes, move on.

[meekGee]

Man, I am sorry but I have to dis you. You want an intelligent conversation with your car? I've come to respect you too much to accept this. Good prompts on where to turn, but conversation...?

As for the phone people at the bank... I needed to find out the source of a cashiers check received by me that originated from a credit union of which I am not a member. All phone numbers let to a commercial call center that could only answer routine questions like "what is my balance?" The best of AI can't compensate for poor business practices.

Of course.

Most of the frustration with automation is just how dumb it is.

Most of the attraction with using chat while engineering is that there's context, and measure, and calibration.  There's even a level of independence in that things come up that I didn't directly ask for.

I don't want to phrase queries exactly the way Google maps expects them, and then have to filter out 90% of garbage results including shops that are behind me, or will be closed, etc. I want to have a natural language conversation about it.

Or say that I'm out of XXX and YYY and have the car figure out how to best get it.  But if I'm on the East coast, it won't try to send me to a Home Depot in Utah because it's on the way.

Basically I want to talk to it like it was a cognizant human being. With wheels and a motor.

For example, it sees all around it, right?  So if an interesting truck drives by, I want to be able to ask WTF was that, and then be able to pull on that string.

I want to be able to comment on an odd looking cloud, or a plane that caught my eye...

Basically the kinda conversation I'd have with a person, except of course colored by the limitations and potential of an AI.

Yup my plastic pal who's fun to be with.

OMG. You want a relationship with your car? I think there might be a generational disjuncture here. I just want a car to run straight, stop straight and keep the water out.

Yeah, old body, young spirit, the secret to my success... 

In seriousness, it's not a relationship in the sense of a human one, but it's interacting with a computer in a human like way, where I don't have to slow myself down and adjust to its dumbness.

And if a chat with your car makes you uncomfortable, think of it as your personal copilot that happens to inhabit the car when you're in it, but goes with you when you get out.

So minus the copilot, we're back to run smooth, turn aharp, and keep the water on the right side of the windshield...

Easier?


(BTW your example with the call center - those piss poor systems are horrible, but are also miles from what good AI can do, even today. Personal chatGPT isn't like that at all.)

We are in such totally different wavelengths. First thing I do with a computer is shut off the sound, yank out as much cruft as I can, disable most everything in startup then only plug into the internet as needed.  If my car started talking to me I'd pull over at a pawn shop, buy a gun and shoot it.


I'm in the market for an SUV. First entry in the decision tree is: Does it have a modem? If yes, move on.

It's the times, man.   They are a-changin'.

I don't auto-welcome any new fad, but some stuff (like TV, or AI) is actually important.  (Did you hate the radio with pictures too when it came out?  How about talkies?)

Also, Merry Christmas!

[OTV Booster]

Man, I am sorry but I have to dis you. You want an intelligent conversation with your car? I've come to respect you too much to accept this. Good prompts on where to turn, but conversation...?

As for the phone people at the bank... I needed to find out the source of a cashiers check received by me that originated from a credit union of which I am not a member. All phone numbers let to a commercial call center that could only answer routine questions like "what is my balance?" The best of AI can't compensate for poor business practices.

Of course.

Most of the frustration with automation is just how dumb it is.

Most of the attraction with using chat while engineering is that there's context, and measure, and calibration.  There's even a level of independence in that things come up that I didn't directly ask for.

I don't want to phrase queries exactly the way Google maps expects them, and then have to filter out 90% of garbage results including shops that are behind me, or will be closed, etc. I want to have a natural language conversation about it.

Or say that I'm out of XXX and YYY and have the car figure out how to best get it.  But if I'm on the East coast, it won't try to send me to a Home Depot in Utah because it's on the way.

Basically I want to talk to it like it was a cognizant human being. With wheels and a motor.

For example, it sees all around it, right?  So if an interesting truck drives by, I want to be able to ask WTF was that, and then be able to pull on that string.

I want to be able to comment on an odd looking cloud, or a plane that caught my eye...

Basically the kinda conversation I'd have with a person, except of course colored by the limitations and potential of an AI.

Yup my plastic pal who's fun to be with.

OMG. You want a relationship with your car? I think there might be a generational disjuncture here. I just want a car to run straight, stop straight and keep the water out.

Yeah, old body, young spirit, the secret to my success... 

In seriousness, it's not a relationship in the sense of a human one, but it's interacting with a computer in a human like way, where I don't have to slow myself down and adjust to its dumbness.

And if a chat with your car makes you uncomfortable, think of it as your personal copilot that happens to inhabit the car when you're in it, but goes with you when you get out.

So minus the copilot, we're back to run smooth, turn aharp, and keep the water on the right side of the windshield...

Easier?


(BTW your example with the call center - those piss poor systems are horrible, but are also miles from what good AI can do, even today. Personal chatGPT isn't like that at all.)

We are in such totally different wavelengths. First thing I do with a computer is shut off the sound, yank out as much cruft as I can, disable most everything in startup then only plug into the internet as needed.  If my car started talking to me I'd pull over at a pawn shop, buy a gun and shoot it.


I'm in the market for an SUV. First entry in the decision tree is: Does it have a modem? If yes, move on.

It's the times, man.   They are a-changin'.

I don't auto-welcome any new fad, but some stuff (like TV, or AI) is actually important.  (Did you hate the radio with pictures too when it came out?  How about talkies?)

Also, Merry Christmas!

Oh yeah, and another thing. Don't wan't that newfangled automatic spark advance neither. That's just for the wimps that can't man up enough to handle a crank starter.


And a belated Merry Christmas back atcha. 

[Twark_Main]

OMG. You want a relationship with your car? I think there might be a generational disjuncture here. I just want a car to run straight, stop straight and keep the water out.

I think there's a big difference between "AI girlfriend" and "the computer voice from Star Trek."

The latter is what is meekGee is actually describing, but you're making it out like he's describing the former. Cheap shot and shallow dismissal. 

[OTV Booster]

OMG. You want a relationship with your car? I think there might be a generational disjuncture here. I just want a car to run straight, stop straight and keep the water out.

I think there's a big difference between "AI girlfriend" and "the computer voice from Star Trek."

The latter is what is meekGee is actually describing, but you're making it out like he's describing the former. Cheap shot and shallow dismissal. 

I have to bow to your expertise in cheap shots and shallow dismissals.

[meekGee]

Hey all's fair in love and AI...

I do have a good story about that btw, my wife thinks it's hilarious (yes Nomadd, we got married, did I tell you?) but it might be a taaaad off topic...

Let's stick with my engineering buddy who's fun to be with.

My .02: Demand for AI, and more specifically demand for AI compute power, will continue to rise at crazy rates.

Hardly anyone today really uses it, the applications are just barely exploring what's feasible, and right now nobody knows how to solve the scalability problem.

[Twark_Main]

OMG. You want a relationship with your car? I think there might be a generational disjuncture here. I just want a car to run straight, stop straight and keep the water out.

I think there's a big difference between "AI girlfriend" and "the computer voice from Star Trek."

The latter is what is meekGee is actually describing, but you're making it out like he's describing the former. Cheap shot and shallow dismissal. 

I have to bow to your expertise in cheap shots and shallow dismissals.

I accept this as your clear admission of guilt, and your complete retraction of the quoted statement as invalid and specious. 

[JaimeZX]

I'm in the market for an SUV. First entry in the decision tree is: Does it have a modem? If yes, move on.

Sadly you'll be looking for a *long* time.  Almost every new vehicle these days "phones home" regularly.

[meekGee]

I'm in the market for an SUV. First entry in the decision tree is: Does it have a modem? If yes, move on.

Sadly you'll be looking for a *long* time.  Almost every new vehicle these days "phones home" regularly.

I have to say tho, while I really like me my AI, I also want a doomsday-ready car that runs even when the Internet is down, service is overdue, and the left front camera cover is dirty...