OP
- Thread Starter
- #2,321
They guy is overhyping this but the technology is still impressive if it proves reliable and producible.
-
Welcome to ASR. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!
Master AI (Artificial Intelligence) Discussion/News Thread
- Thread starter amirm
- Start date
digitalfrost
Major Contributor
I've been testing the GPT 5.6 models since they came out and I have to say I'm impressed. 5.6 Sol especially will just execute a task and not give up. I gave it stuff to do and it just keeps working on it for 2 hours or longer without stopping or asking stupid questions. This means you can start something before going to bed and you know when you wake up it will be done.
How many tokens does it use in 2 hours?
digitalfrost
Major Contributor
It depends heavily but for example one of my recent goals was "Goal completed in approximately 1 hour 47 minutes, using 1,201,092 tokens."
Sol Ultra sucks usage like crazy even on the pro Plan I would not recommend it unless you have a very hard problem or need absolute best quality. I did a code audit with Sol Ultra for a project of mine and while the results were stellar after working for 20mins it used up 2/3rds of my 5 hour usage window.
edit: Another example that just finished "Goal completed in 19 minutes 24 seconds, using 422,077 tokens." so you see...it really depends.
Last edited:
Director of many great films, Christopher Nolan also has some interesting things to say about AI. As in, it's basically womp womp
www.theguardian.com
Christopher Nolan says people ‘disdain’ AI and the idea it will replace humans is ‘nonsense’
Odyssey director addresses industry fears over artificial intelligence and says rightwing criticism of Lupita Nyong’o as Helen of Troy is ‘irrelevant’
www.theguardian.com
That young people turning away from AI generated content (AI slop) is understandable, but I suspect it is often driven by deeper instincts than just a dislike of the technology itself. What i already stated in a previous reply is that human existence has always been shaped by linear time, biological survival and entropy. Because of that we compete because resources, time and lifespan are limited. From that perspective, it is not surprising that a technology appearing to challenge human uniqueness can trigger an almost existential response.Director of many great films, Christopher Nolan also has some interesting things to say about AI. As in, it's basically womp womp
Christopher Nolan says people ‘disdain’ AI and the idea it will replace humans is ‘nonsense’
Odyssey director addresses industry fears over artificial intelligence and says rightwing criticism of Lupita Nyong’o as Helen of Troy is ‘irrelevant’
IMO the problem is not AI itself, it is more often a lack of understanding of how to use it properly. There is a fundamental difference between simply asking AI to produce something and actually directing it with intention. The latter still requires human creativity experience/knowledge, taste, domain expertise and Intelligence.
Personal (subjectively) what I see around me tells a different story. The generation of programmers after the baby boomers Generation X 46 to 61 year old ( who represent most of the people i know) appear to use AI rapidly. They invest in payed AI tools, build agents, learn prompt engineering an it looks they reposition themselves within the IT world. Not because they think AI will replace them, but because they understand that the people who know how to guide, challenge and integrate AI will likely remain the most relevant.
The million dollar question is how long that advantage will last. But one thing already seems clear, AI does not replace knowledge; it amplifies the value/inteligence of people who know how to apply it. It seems the real power lies in the synergy between human judgment and machine capability.
Or as the old saying goes: in the land of the blind, the one eyed man is king. Right now, the people who understand both worlds may just have the clearest vision
Last edited:
- Joined
- Jan 20, 2019
- Messages
- 621
- Likes
- 406
what is a token? what kind of tasks are you having the thing work on?
I use AI mostly for writing code and some general knowledge questions. Tokens are how AI keeps track of it's time for billing purposes. Fancy frontier models cost a lot more than older models. A poster mentioned he had a frontier model running for almost 2 hours on a programing task which in my experience could use up hundreds of dollars worth of tokens so I was curious. The calculation is not as straightforward as "cost per token" so a lot "depends" on the applications. See OpenAI pricing https://developers.openai.com/api/docs/pricing?latest-pricing=priority
Petrushka
Major Contributor
- Joined
- Jun 29, 2024
- Messages
- 1,784
- Likes
- 2,081
Does anyone remember Offshore Power Systems, a failed effort to move nuclear power to artificial reefs, to overcome meltdown fears?
I knew an engineer who was denied permission to run a program because it would take 0.8 seconds of supercomputer time.
Now we are making AI cat videos. Semiconductor prices have spiked because demand has temporarily outpaced supply, but that will change. Compute will move to solar powered satellites. Welcome to SkyNet.
Was the supercomputer more or less powerful that a current day cell phone?
- Joined
- Dec 12, 2019
- Messages
- 17,731
- Likes
- 40,697
So -- presumably somewhere in space-time -- someone is asking some suitably capable AI agent to solve the problem of controlled, sustainable nuclear fusion.
That would solve one problem.
That would solve one problem.
If I was king of the world I would not allow anyone to sell AI services until they first used it to cure cancer.So -- presumably somewhere in space-time -- someone is asking some suitably capable AI agent to solve the problem of controlled, sustainable nuclear fusion.
That would solve one problem.
It is an almost solved problem, that usage of regenerative energy (wind, water and solar), in combination with storage (battery, hydrogen ...) will make any different energy generation technology obsolete in the near future ... especially accounting the " within the next 30 years" mentioned according nuclear fusion, beginning from 50th of last century.So -- presumably somewhere in space-time -- someone is asking some suitably capable AI agent to solve the problem of controlled, sustainable nuclear fusion.
That would solve one problem.
Petrushka
Major Contributor
- Joined
- Jun 29, 2024
- Messages
- 1,784
- Likes
- 2,081
Given fifty years we could local the largest energy users near sources of solar energy, but there will always be a need for backup.
Solar panels can be destroyed by storms and rendered useless by cloud coverage. My personal experience is that clouds reduce energy capture by 95 percent.
The paradox of energy transition like in China,: more solar, wind, batteries and EVs mean less fossil fuel burning and cleaner air. But cleaner air comes with an unexpected side effecy. Fewer aerosols mean removing a temporary “sun shield.” Scientists call this the clean air warming penalty. Cleaner skies can actually cause a temporary local warming effect. It’s not an argument for pollution, just a fascinating reminder of how complex the climate system really is.
Climate responses in China to domestic and foreign aerosol changes due to clean air actions during 2013–2019 - npj Climate and Atmospheric Science
In recent years, to improve air quality, significant efforts have been made to reduce regional aerosols including China, Europe and North America, which have potential impacts on climate. In this study, fast and slow climate responses in China to changes in domestic and foreign anthropogenic...
www.nature.com
- Joined
- Apr 24, 2022
- Messages
- 2,076
- Likes
- 2,777
The Wall Street Journal reports A.I. companies are seeing an increase in threats of violence and are spending more money on security to protect their employees. (gift link below)
Sample size of 1. I no longer initiate a conversation about A.I. with select friends and relatives in the workforce. It may not be welcomed.
Sample size of 1. I no longer initiate a conversation about A.I. with select friends and relatives in the workforce. It may not be welcomed.
OP
- Thread Starter
- #2,337
Hadn't thought about that development. Minting a large number of millionaires with their IPOs is only going to make this worse.
This part of the WSJ article really stood out for me:
"Americans concerned about AI outnumber those who aren’t by more than a 4 to 1 margin, according to a March survey of about 1,400 U.S. adults by Quinnipiac University. A growing share of respondents to that survey—55%—said they believed AI was doing more harm than good."
This is a crisis if there ever was one. These guys need a serious plan to show value of AI beyond putting folks out of a job.
This part of the WSJ article really stood out for me:
"Americans concerned about AI outnumber those who aren’t by more than a 4 to 1 margin, according to a March survey of about 1,400 U.S. adults by Quinnipiac University. A growing share of respondents to that survey—55%—said they believed AI was doing more harm than good."
This is a crisis if there ever was one. These guys need a serious plan to show value of AI beyond putting folks out of a job.
- Joined
- Apr 24, 2022
- Messages
- 2,076
- Likes
- 2,777
Agreed.
With the perceptions about A.I., imagine what an A.I. company now has to do to create the perception amongst its employees that it is safe to work on-site. I can imagine a skit on Saturday Night Live where the establishing shot of a large tech campus is pictured to resemble 'The Green Zone' in Baghdad, Iraq.
In the early days of the British industrial revolution, when the labour intensive process's of making cotton and woollen cloth were being mechanised, bands of workers would go around destroying the new machinery that were putting them out of work. They were known as luddites - which is where we get the modern, usually pejorative term, describing someone opposed to technology.
Will we see a new breed of 'luddites' (In the original meaning) attempting to disrupt the operation of AI? Perhaps by hacking, or cutting off energy, or even bombing data centres?
I think we might if the mass unemployment being described in some circles actually comes to pass - without some massive sociological change to avoid impoverishment of people replaced by AI.
Enthusiastic Amateur
Member
Bruce Schneier is one of the world's most respected and eloquent cryptography/security experts. A deep thinker who maintains a very sensible blog, he has recently been focusing on the effects of AI on democracy. Just a taste:
www.schneier.com
AI Data Centers and the Concentration of Wealth - Schneier on Security
This essay was written with Nathan E. Sanders, and originally appeared in The Guardian. Opposition to AI data centers has emerged as a primary theme in US politics, one that—surprisingly—doesn’t fall along party lines. We applaud people coming together for constructive debate on any issue, and...
www.schneier.com
Similar threads
