"Blasted blubber beyond all believable bounds".
Now that's the type of fine journalism missing in todays reporting.
Do you handle change well? How much change has our current generations REALLY been exposed to?
- Thread starter SammyChevelleTypeS3
- Start date
"Blasted blubber beyond all believable bounds".
Now that's the type of fine journalism missing in todays reporting.
No, but I've seen bits where they do have robots walking around. Might be sooner than we think--also, the fine art of following a checklist to determine a problem--then using knowledge of the system to go above and beyond any checklist, I suspect AI might wind up being better at doing debug than people.
I'm fascinated by the way humanoid robots have developed in the last few years. They have got the balance issue sorted but it comes at a cost of continuous small movements which uses energy and that means battery life is low at typically 2 hours. I suspect the power source will be a bigger nut to crack than the robot itself.
A robot swapping it's own battery.
<iframe width="560" height="315" src="" title="YouTube video player" frameborder="0" allow="accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture; web-share" referrerpolicy="strict-origin-when-cross-origin" allowfullscreen></iframe>
I have yet to see any problem solving beyond programming. Like what if the plans weren't followed? Would the robot be able to think like a lazy human or a human that doesn't follow plans? Would be interesting if they could.
Reverse engineering AI between competing models from different countries and corporations.
That's the thing--but how many ordinary folk can think outside of what they've been told how to? At least with AI, once taught how to take wild guesses when all else has failed, perhaps it will figure out more complex problems by being more willing to "guess". Not a simple thing to learn--human or otherwise.
I do think a few of the things that happens in the last 10 or so years are not all about AI. The automation / robotics advances are more about higher level of applications and automations due to faster and more compact power electronics instead of what we called "generative AI" today. Sure we no longer just use a control loop that runs slowly but we have a lot of new improvements that would have come along with or without generative AI (like how Boston Dynamics came out with their robots without large language model). The tech was there as a form of DSP doing the math instead of GPU, but GPU became economical because of gaming then crypto mining. Had they need a 1990s super computer to run LLM we wouldn't have today's AI boom.
Also about self swapping battery. It is actually not really a challenge if you see how we have data center redundant power supply swapping and fail over, or things like hot swap of hard drive. It might be cool but it is not actually that much harder, and not really what we called AI. It is now ready because the popularity of smart phone and cloud computing made these electronics economical for the mass.
About today's AI, I think they are showing a lot of promise but we likely will see something unforeseen beyond today's agent. The training is all statistical today but we probably would hit a bottleneck unless they can break it down into multiple model connected to each others and use deterministic "algorithm" well laid out by mathematicians and scientists in between. Today's model can easily be wrong if you train from public opinion, like how a teenager can be influenced to do dumb things. AI need to one day move beyond that. Maybe they need authority laying out the deterministic rules, maybe they need the authority to run experiments in a lab to fill in its own blank, maybe it needs to do something we have not seen before, who knows. They won't be able to destroy the world unless we give them the authority though, just like how a human can't do that unless we give him the authority.
Also about self swapping battery. It is actually not really a challenge if you see how we have data center redundant power supply swapping and fail over, or things like hot swap of hard drive. It might be cool but it is not actually that much harder, and not really what we called AI. It is now ready because the popularity of smart phone and cloud computing made these electronics economical for the mass.
About today's AI, I think they are showing a lot of promise but we likely will see something unforeseen beyond today's agent. The training is all statistical today but we probably would hit a bottleneck unless they can break it down into multiple model connected to each others and use deterministic "algorithm" well laid out by mathematicians and scientists in between. Today's model can easily be wrong if you train from public opinion, like how a teenager can be influenced to do dumb things. AI need to one day move beyond that. Maybe they need authority laying out the deterministic rules, maybe they need the authority to run experiments in a lab to fill in its own blank, maybe it needs to do something we have not seen before, who knows. They won't be able to destroy the world unless we give them the authority though, just like how a human can't do that unless we give him the authority.
For that to happen AI need to compete with each other and evolve, learn from each other from mistakes, etc. So far we haven't seen AI models competition between large companies yet.
DARPA probably should do some of these challenges and allow other countries to compete with us as well. They have sponsored a lot of those competitions back then and breed the hybrid we drive today (those competition back in the mid-late 90s), humanoid robots we have today were probably from the early 2000s DARPA stuff, and self driving as well back then.
Good point. At least robots follow the instrutions and rules they have been given; humans not so much. A huge detriment to self driving cars is human error; humans don't follow the rules of the road which AI has to accomodate for.
AI can learn from its mistakes via the feedback loop. There are also advanced techniques that help by reframing failed attempts as successes, allowing the AI to learn faster from its mistakes.
Yeah and is the reason that most everyone now is stupid.
Its called information overload, everyone's minds are so jumbled up with garbage that there is no room for thinking.
Yeah anything I learned 40 years ago is outdated, but not the stuff I learned 55 years ago.
4WD
$50 site donor 2025
Maintenance crew at the AI facility
- Joined
- Dec 7, 2022
- Messages
- 226
- Joined
- Dec 7, 2022
- Messages
- 226
Seems like the common denominator is drugs. Maybe it's time to wise up and get rid of drugs and alcohol.
Come watch the documentaries with me. They are really good, they got Arnold to play the sentient AI.
Funny how a thread about change, inevitably turns into AI thread. Is the the only change people see?
This is good but giving it some more thought I realize that the younger generations can in many cases look back to their elders into industries that still require labor (and some well paid labor at that) where technology hasn't and doesnt appear to be able to replace that portion of the workforce.
Pretty much all trades that are mobile and require hands on work. Plumbers, Electricians, Roofers, Climate control technicians, engineers, which also includes the very buildings housing trillions of dollars worth of network/computer systems. Yeah, those systems and AI are dependent on humans keeping them cool. (you can also own these companies and do not need to do the physical work) Lets not forget about many much more higher education levels such as doctors/surgeons.
I was born in 1953 and have seen a little bit of change. I remember my mother's parents lighting up the wood burning kitchen stove in the morning to both heat the house and cook breakfast. I remember downtown Seattle and Tacoma being great places to shop at Christmas, or any other time. I remember our first color TV, first microwave, first home computer, windmills on farms, being on the beach watching the missile launch for the first man in space, figuring this was it in October 1962, and figuring Vietnam was a likely possibility once I graduated from high school, etc.
Change is the one constant in life.
Change is the one constant in life.
To be fair, that's what the media is cramming down our throats right now.
