I ask AI, chatgpt 4 how the world will change

Avatar image for blaznwiipspman1
blaznwiipspman1

15812

Forum Posts

0

Wiki Points

0

Followers

Reviews: 0

User Lists: 0

#1  Edited By blaznwiipspman1
Member since 2007 • 15812 Posts

So I am paying for the latest model of chatgpt 4 and honestly its amazing. The thing is so powerful, it can wipe out almost every single white collar job, including doctors, lawyers, software developers, etc etc. The capability is that good. I asked chatgpt4 a question:

Q: "In a world where most white collar jobs are replaced or significantly scaled back by AI, how will policy change? For ex, would we have UBI for every citizen to reap the benefits of the productivity, or would the rich hord all the wealth from the new AI tech?"

Chatgpt4:

"In a world where most white collar jobs are replaced or significantly scaled back by AI, several policy changes could occur to address the resulting economic and social consequences. It is important to note that these changes would depend on the political climate, cultural values, and economic systems in place. Some potential policy changes might include:

  1. Universal Basic Income (UBI): To address the widespread job displacement and income inequality that could result from AI automation, governments might implement a Universal Basic Income. This policy would provide every citizen with a regular, unconditional sum of money to ensure a basic standard of living, irrespective of their employment status. UBI could help redistribute the wealth generated by AI productivity and allow people to pursue meaningful activities outside the traditional labor market.
  2. Wealth Redistribution: As AI technology becomes more advanced and accessible, the wealth generated by this technology could become highly concentrated among a few companies and individuals. To prevent this, governments could implement policies like progressive taxation, wealth taxes, or even a robot tax, where companies are taxed based on the degree of automation they employ.
  3. Education and Retraining Programs: With the decline of white collar jobs, governments might invest more in education and retraining programs to help people transition into new careers or industries. This could include providing incentives for lifelong learning, upskilling, or reskilling initiatives, focusing on sectors that are less likely to be automated.
  4. Encouraging Entrepreneurship and Innovation: Policy changes could focus on fostering innovation and entrepreneurship to create new industries and jobs. Governments might offer incentives, grants, or tax breaks to start-ups and small businesses, while also investing in research and development.
  5. Social Safety Net Expansion: As job security becomes less certain in a highly automated world, governments might strengthen social safety nets by expanding access to healthcare, housing, and other essential services, ensuring that citizens can maintain a decent standard of living even if they experience job displacement.
  6. Labor Rights and Regulation: As the nature of work changes, labor policies might need to be updated to protect workers' rights in the face of increased automation. This could include implementing policies that promote job-sharing, flexible work arrangements, or advocating for a shorter work week to distribute work more evenly across the population.

In conclusion, there are many potential policy changes that could be implemented to address the challenges posed by widespread AI automation in white collar jobs. The specific changes would depend on the political climate, cultural values, and economic systems in place, but the key would be to ensure that the wealth generated by AI technology is distributed fairly and that citizens have the opportunity to pursue fulfilling lives."

What are your thoughts? Are we headed towards more socialism? If you have any questions you want to ask chatgpt4, let me know. I do have a limit of 25 messages though.

Avatar image for blaznwiipspman1
blaznwiipspman1

15812

Forum Posts

0

Wiki Points

0

Followers

Reviews: 0

User Lists: 0

#3 blaznwiipspman1
Member since 2007 • 15812 Posts

@girlusocrazy: yeah you should. It's worth it to get chat gpt4 even with the limit in messages. This thing is crazy awesome.

Avatar image for sargentd
SargentD

6811

Forum Posts

0

Wiki Points

0

Followers

Reviews: 0

User Lists: 5

#4 SargentD
Member since 2020 • 6811 Posts

Ask it if AI can have morals. If it can decipher between right and wrong on a personal level.

Avatar image for blaznwiipspman1
blaznwiipspman1

15812

Forum Posts

0

Wiki Points

0

Followers

Reviews: 0

User Lists: 0

#5  Edited By blaznwiipspman1
Member since 2007 • 15812 Posts

@sargentd said:

Ask it if AI can have morals. If it can decipher between right and wrong on a personal level.

Q: do you think AI can have morals? Will it be able to tell between right and wrong?

chatgpt4 tldr version:

"AI systems do not inherently have morals, but they can be designed to mimic moral reasoning based on ethical guidelines set by their creators. Their understanding of right and wrong is limited to their training data and the rules provided by developers. Ensuring ethically sound AI requires addressing biases in data, designing algorithms with ethical principles, and collaborating with experts in ethics and social sciences."

chatgpt4 long version:

"AI systems, like the one you are currently interacting with, do not have feelings, consciousness, or an inherent sense of morality. They are essentially advanced tools that process and generate text based on the data they have been trained on.

However, AI systems can be designed to mimic moral reasoning and follow ethical guidelines set by their creators. They can be programmed to make decisions based on certain ethical principles, like utilitarianism or deontological ethics. These principles can be encoded as rules, heuristics, or objective functions that guide the AI's decision-making process.

It is crucial to note that an AI's understanding of right and wrong is limited to the data it has been trained on and the rules and objectives provided by its creators. It may still make morally questionable or biased decisions, especially when faced with novel situations that were not adequately covered in its training data or when the data itself contains biases.

To develop AI systems that are more capable of making ethically sound decisions, researchers and developers need to invest time and effort into understanding and addressing the ethical implications of AI. This involves studying the potential biases in training data, designing algorithms that can reason about ethical principles, and collaborating with experts in ethics and social sciences."

Avatar image for blaznwiipspman1
blaznwiipspman1

15812

Forum Posts

0

Wiki Points

0

Followers

Reviews: 0

User Lists: 0

#6 blaznwiipspman1
Member since 2007 • 15812 Posts

@sargentd: personally I think this is a rabbit hole nobody is going to like, in the future. Especially conservatives. Most conservative values clash with the mainstream liberal ones, so its going to be interesting to see how AI is programmed.

Avatar image for blaznwiipspman1
blaznwiipspman1

15812

Forum Posts

0

Wiki Points

0

Followers

Reviews: 0

User Lists: 0

#8  Edited By blaznwiipspman1
Member since 2007 • 15812 Posts

@girlusocrazy: my man, even though I have a feeling this AI will wipe out my job in tech, I'm still super excited over it. Losing your job, that's just life 🤣. This thing is absolutely incredible though.

Chatgpt3.5 has alot of errors, but version 4 is a game changer, it's far more reliable. I reccomend you give it a shot, even for a month. It will blow you away.

Avatar image for sargentd
SargentD

6811

Forum Posts

0

Wiki Points

0

Followers

Reviews: 0

User Lists: 5

#9 SargentD
Member since 2020 • 6811 Posts

@blaznwiipspman1 said:

@sargentd: personally I think this is a rabbit hole nobody is going to like, in the future. Especially conservatives. Most conservative values clash with the mainstream liberal ones, so its going to be interesting to see how AI is programmed.

I get what your saying. I mean your right but at the end of the day it's going to come down to if you value what the bot/ai says, or you don't.

Personally I could care less what it says, it's programmed by people, and people will program it how they see fit. It's bias will come from gathering information written by people with their own bias.

Personally I don't trust the bots. What scares me is the people that will put too much trust into it. People that will put heavy worth into what these bots say thinking it's more "trustworthy". The AI doesn't worry me as much as the people that will treat the AI as some arbiter of truth.

As far as the AI taking over and taking away jobs.

Yeah it will, it's funny too that 5 years ago people were saying AI will be used to take away tedious work and allow humanity to do more life fulfilling work like the arts for example. Then the A.I started doing that too lmao.

Hands on labor will be the last type of labor this AI will have to compete with humans on. It's one thing creating information based A.I, much more difficult to have Ai build you a house!

I can only hope that one day in the future I can join a rebellion to destroy the AI completely to save humanity from A.I tyrants.

Avatar image for blaznwiipspman1
blaznwiipspman1

15812

Forum Posts

0

Wiki Points

0

Followers

Reviews: 0

User Lists: 0

#11 blaznwiipspman1
Member since 2007 • 15812 Posts

@girlusocrazy: lmfao 🤣 you gave it a shot, that's all you can say.

Avatar image for blaznwiipspman1
blaznwiipspman1

15812

Forum Posts

0

Wiki Points

0

Followers

Reviews: 0

User Lists: 0

#12  Edited By blaznwiipspman1
Member since 2007 • 15812 Posts

@sargentd: yep, people never expected the AI to do creative jobs, and now they're all shaking. As a tech worker, I can see the writing on the wall. My job or role will definitely change in a big way.

You're right about the construction and trade jobs being the very last to get automated. It's just very difficult to make a robotic body that can interact with the world perfectly. Plus the technology to build it is very expensive, and it wouldn't be able to do as well as humans. Even still, I expect them to automate this field too at some point, maybe 20 or 30 years.

🤣 I hope it doesn't get to that point where we have a robot war, because if it does, that means the robots have become sentient and that is likely not a battle we would win.

Avatar image for tjandmia
tjandmia

3496

Forum Posts

0

Wiki Points

0

Followers

Reviews: 0

User Lists: 5

#14  Edited By tjandmia
Member since 2017 • 3496 Posts

So what's this, are Americans finally realizing that wealth inequality is a big problem, and that they can't keep prioritizing the whims of the wealthy over the needs of everyone else? It took AI for them to understand this?

Avatar image for palasta
palasta

1204

Forum Posts

0

Wiki Points

0

Followers

Reviews: 0

User Lists: 5

#15 palasta
Member since 2017 • 1204 Posts

Yes, it can c&p quite well. Amazing.

When i put the AI to the test, it failed miserably. The problem wasn't even that it couldn't complete the task. The suggestions it gave often didn't match the given information, contrasted them, or/and it included given details in the descriptions to match the critera. Using google would be more effective.