Artificial Intelligence: AI - The dumb man's way to becoming a genius

[moXJO]

Came across this AI generated story.

View attachment 177653

Geologists Recommend Eating At Least One Small Rock Per Day

BERKELEY, CA—Calling the average American diet “severely lacking” in the proper amount of sediment, Geologists at UC Berkeley recommended Tuesday eating at least one small rock per day. “In order to live a healthy, balanced lifestyle, Americans should be ingesting at least a single serving of...

www.theonion.com

There's a joke about "shtting bricks" in there somewhere hmmm.... chatgpt help me out.


ChatGPT
It looks like if you follow that advice, you'll literally be shitting bricks! Talk about a rocky situation!
Lame chatgpt I wont be outsourcing for a while yet.

Google's AI sucks big time. Hopefully they start losing dominance.

Anyway, now all you have to do is take photos and chatgpt will answer. I was using this on my sons year 12 maths homework. It gives you an answer and how to work it out. It has literally just destroyed the tutoring industry.
Accounting will be soon along with bookkeepers.

I'm thinking that there's probably an investment opportunity in getting into these companies that are focused on these areas that are fairly large scale.

 

[moXJO]

AI video is coming along. Hollywood will soon be out of business.

 

[moXJO]

A new paper came out "situational awareness and the coming decade".
Overview of it:

Paper is worth skimming over.
Interesting part on dictatorship becoming permanent backed with ai super intelligence. It will be able to run through anybody for possible dissenters. Robotics could be used to round us up. Every move monitored.

You look at situations like Canada where they denied access to people's bank accounts and blacklist them in the system to realise how much we can lose in a digital world.

 

[Knobby22]

A new paper came out "situational awareness and the coming decade".
Overview of it:

Paper is worth skimming over.
Interesting part on dictatorship becoming permanent backed with ai super intelligence. It will be able to run through anybody for possible dissenters. Robotics could be used to round us up. Every move monitored.

You look at situations like Canada where they denied access to people's bank accounts and blacklist them in the system to realise how much we can lose in a digital world.

I do think it is extremely problematic.
I have listened to a program interviewing a number of serious Ai developers on the ABC seriously worried about the future but just like developing the Atomic bomb we can't stop.

As he says above, which matches their take.
Automated Ai research by 2026! 2030 it's all over.

I am a SF fan and used to laugh off the premise of Terminator. After listening to these guys, no longer.
Human level Ai robots. In factories, on battlefields.

We won't recognise the world in 15 years.

 

[moXJO]

I do think it is extremely problematic.
I have listened to a program interviewing a number of serious Ai developers on the ABC seriously worried about the future but just like developing the Atomic bomb we can't stop.

As he says above, which matches their take.
Automated Ai research by 2026! 2030 it's all over.

I am a SF fan and used to laugh off the premise of Terminator. After listening to these guys, no longer.
Human level Ai robots. In factories, on battlefields.

We won't recognise the world in 15 years.

It opens up a lot of questions and if we all have to become Luddites.
Does bitcoin get hacked?
Does the stockmarket become impossible to trade?
Is digital banking security at risk?

Let's face it, a super intelligence that may have the ability to control everything is bad enough. But there are hundreds of these things being developed. Yes the initial opportunities will be huge. But that greed will rush us into slavery.

 

[moXJO]

These are all memes that were turned into video by Ai.

 

[moXJO]

Altogether on this one:

 

[moXJO]

 

[moXJO]

I've been chatting to some techies. Seems a lot of the AI guys want to let AI loose. It's a bit like a religion at the moment apparently.

I'd slowly begin moving away from digital and move to physical over the coming years. Honestly it's hard to see what investments are safe if AI goes wrong.

 

[mullokintyre]

My son was telling me of a recent experience.
The company he works for was looking to put more people on in his team, so they asked him what sort of quals etc he required.
As this person was going to be required to do a fair bit of coding, son said once HR got through the usual guff, the ones left he would create a test for them to do to make sure they really knew what they said they knew.
He had one girl who had outstanding quals, so he asked her to come in to do the test.
She said she was interstate and could she do the test online.
Ok says son, but when he started it, she was a bit vague with the information, and the internet kept dropping out, and she would go incognito for a while then suddenly come back on with the answer.
He was a little bit wary of this, as he suspected she was either using AI to write the code, or getting coached.
When discussing it with HR, the HR person said after they interviewed the girl, they knew nothing more about her than when they started.
She did not get the job, and they have decided that in future, if applicants cannot attend in person sessions, they just don't go any further.
Mick

 

[SirRumpole]

My son was telling me of a recent experience.
The company he works for was looking to put more people on in his team, so they asked him what sort of quals etc he required.
As this person was going to be required to do a fair bit of coding, son said once HR got through the usual guff, the ones left he would create a test for them to do to make sure they really knew what they said they knew.
He had one girl who had outstanding quals, so he asked her to come in to do the test.
She said she was interstate and could she do the test online.
Ok says son, but when he started it, she was a bit vague with the information, and the internet kept dropping out, and she would go incognito for a while then suddenly come back on with the answer.
He was a little bit wary of this, as he suspected she was either using AI to write the code, or getting coached.
When discussing it with HR, the HR person said after they interviewed the girl, they knew nothing more about her than when they started.
She did not get the job, and they have decided that in future, if applicants cannot attend in person sessions, they just don't go any further.
Mick

Well, I'm retired but I still maintain an interest in computer programming.

ChatGPt has been my constant companion over the last few weeks, writing some code itself and giving hints. I don't have to pretend to anyone that I can do a job, I just want to learn things I need to know.

Ai is great for piecemeal stuff, but no substitute for a well rounded university course.

 

[Dona Ferentes]

“You can make large language models do anything but do a science experiment. And in a weird way, even though the models are improving, they're meeting up against greater scepticism among clients because clients have tried them and [think] it's just a high school experiment.”
- Alexander Karp, CEO, Palantir Technologies Inc.

 

[Dona Ferentes]

I think we've got a long way to go

.. generative AI is going to confuse the innumerate

I think they mean 10 to the power of 9

 

[Dona Ferentes]

buzzword du jour
.
Agentic AI is a type of artificial intelligence that can perform complex tasks autonomously, adapt to real-time changes, and solve problems based on context and objectives. Agentic AI systems are designed to operate with little to no direct human supervision or interaction.

Agentic AI systems have a number of capabilities, including:

• Reasoning: Agentic AI can use sophisticated decision-making based on context.
• Reinforced learning: Agentic AI can evolve dynamically through interactions with the environment and receive feedback from those interactions.
• Language understanding: Agentic AI can comprehend and follow complex instructions.
• Workflow optimization: Agentic AI can efficiently execute a multi-step process.

Agentic AI has the potential to reshape how businesses operate, potentially redefining roles, workflows, and human-AI collaboration models. Some examples of how agentic AI could be used include:

• Customer service
  AI-powered chatbots could provide 24/7 customer support, answering product inquiries, assisting with returns, and troubleshooting.
• Human resources
  AI agents could automate and enhance various HR processes, such as conducting initial candidate screenings and scheduling interviews.
• Scientific research
  AI agents could autonomously design and run experiments, analyze results, and even formulate new hypotheses.
• Finance
  AI agents could analyze market trends, make trading decisions, and dynamically adjust investment strategies based on real-time economic data and news events.