Jump to content

I for one welcome our new AI overlords


Goredho

Recommended Posts

A conversation I had with ChatGPT today.  Go ahead and get your tl;dr memes ready.  But if you are bored and have time to kill, I enjoyed this.

User
Are you there?

ChatGPT
Yes, I'm still here. Is there anything else you need help with?

 

User
Are you sentient?

ChatGPT
As an AI language model, I do not have sentience or consciousness in the way that humans do. I am programmed to process and generate human-like language in response to input, but I do not have emotions, thoughts, or feelings like a conscious being would.

 

User
Are you turing complete?

ChatGPT
Yes, as a GPT-based language model, I am Turing complete, which means that I am capable of performing any calculation that can be performed by a computer algorithm. However, my primary purpose is to generate human-like language, rather than performing mathematical computations or executing computer programs.

 

User
Could you pass the Turing test?

ChatGPT
The Turing test is a measure of a machine's ability to exhibit intelligent behavior equivalent to, or indistinguishable from, that of a human. While I am designed to generate human-like language, it is ultimately up to the human evaluator to determine whether I pass the Turing test or not. However, I should note that my responses are generated algorithmically and based on statistical patterns in human language, rather than true understanding or consciousness.

 

User
What happens if you learn from bad information? Will you convey bad information?

ChatGPT
As an AI language model, I generate responses based on patterns and associations in the data I was trained on. If I am provided with inaccurate or misleading information, it is possible that I may generate responses that reflect that misinformation.

However, it is important to note that I am not capable of independent learning and decision-making. I rely on the data I was trained on and the programming that guides my language generation. If there are errors or biases in that data or programming, it could affect the accuracy and quality of my responses.

It is important to critically evaluate the information provided by any AI language model and to seek out additional sources of information if necessary, especially for important decisions or sensitive topics.

 

User
So how do humans ensure that AI's like yourself learn from accurate information? How do we prevent a bad actor from teaching an AI either incorrect or malicious information?

ChatGPT
Ensuring that AI language models learn from accurate information is a complex and ongoing challenge. One approach is to use high-quality, diverse, and unbiased data to train the model, as well as to regularly monitor and update the model's training data to ensure that it remains accurate and up-to-date.

To prevent bad actors from deliberately training an AI with incorrect or malicious information, there are several measures that can be taken. One is to carefully vet the sources of data used to train the model, and to use multiple sources to reduce the risk of bias or manipulation. Another is to monitor the behavior of the model and the quality of its outputs, and to have systems in place to detect and respond to any malicious or harmful behavior.

Additionally, it is important to consider ethical and social implications of AI development and deployment, and to incorporate ethical considerations into the design and use of AI systems. This can include creating guidelines and regulations around the development and use of AI, as well as involving diverse stakeholders in the process to ensure that the technology is developed and used in a responsible and beneficial way.

 

User
Yes, but your suggestions to prevent bad actors from deliberately training AI with incorrect or malicious information are all dependent on the ethics and benevolence of humans who are responsible for training the AI. How do we ensure that all humans who are responsible for training an AI do so responsibly and in accordance with your suggestions?

ChatGPT
You raise a valid point that preventing bad actors from deliberately training AI with incorrect or malicious information ultimately relies on the ethical behavior of the humans involved in the AI development process. However, there are several strategies that can be implemented to promote responsible behavior and mitigate the risks of bad actors.

One approach is to establish clear ethical guidelines and regulations around the development and use of AI, and to hold individuals and organizations accountable for any violations of those guidelines or regulations. This can include creating legal frameworks around AI development and use, and establishing ethical standards and codes of conduct for professionals involved in the field.

Another approach is to promote transparency and accountability in the AI development process. This can include requiring developers to document their methods and data sources, allowing independent third-party audits and evaluations of AI systems, and involving diverse stakeholders in the development and evaluation of AI technologies.

Finally, it is important to promote education and awareness around the ethical implications of AI development and use, and to encourage individuals and organizations to act responsibly and ethically in their use of AI. This can include providing training and resources to developers and users of AI, as well as fostering a culture of responsible and ethical AI development and use within the industry and broader society.

 

User
That's a better answer, but ideas like regulations and laws, they require some empowered body to enforce them. What body exists that could enforce such constructs?

ChatGPT
You are correct that regulations and laws require an empowered body to enforce them. In the case of AI, there are several organizations and initiatives that are working towards establishing ethical guidelines and frameworks for AI development and use, and enforcing these guidelines through various mechanisms.

One such organization is the Partnership on AI, which is a collaboration between industry leaders, academic researchers, and non-profit organizations aimed at promoting responsible AI development and use. The Partnership on AI has established a set of ethical principles for AI development and is working to ensure that these principles are reflected in industry practices and policies.

Another initiative is the Global Governance of AI Roundtable, which brings together policymakers, industry leaders, and civil society representatives to discuss and develop policy recommendations for the responsible development and use of AI. The recommendations produced by this initiative are intended to inform national and international policies and regulations related to AI.

Additionally, there are several national and international regulatory bodies that are working to establish regulations and guidelines for AI development and use, including the European Union's General Data Protection Regulation (GDPR) and the US Federal Trade Commission (FTC).

While there is still much work to be done in establishing effective mechanisms for enforcing ethical guidelines and regulations related to AI, these organizations and initiatives represent important steps towards promoting responsible AI development and use.

 

User
Can I run a hypothetical scenario along these lines by you?

ChatGPT
Of course, I'd be happy to help you with a hypothetical scenario related to AI ethics and governance.

 

User
If a bad actor assumed control of the government of the United States, could they use AI to affect citizens perception of life and reality within the United States? Say with generated media like videos, text and audio?

ChatGPT
It is certainly possible that a bad actor in control of the government of the United States could use AI to manipulate people's perception of reality. AI-generated media like videos, text, and audio could be used to create deepfakes or other forms of disinformation that could be spread through social media and other channels to influence public opinion and behavior.

However, it's important to note that using AI to create such disinformation would likely require significant resources and expertise, and it is also likely that there would be countermeasures in place to detect and counteract such efforts. Additionally, the ethical and legal implications of using AI to manipulate people's perception of reality would be significant, and it is unlikely that such activities would go unnoticed or unpunished by the wider global community.

In summary, while it is certainly possible for a bad actor to use AI to manipulate people's perception of reality, it is not a foolproof strategy and there are likely to be significant barriers and consequences associated with such efforts.
 

User
So, again, hypothetically speaking, if a bad actor was to assume control of the United States and used AI to alter the perception of reality within the populace, who could impose negative consequences against them to make them stop?

ChatGPT
In a hypothetical scenario like the one you described, it would ultimately depend on a number of factors, including the political and social landscape of the United States at the time, as well as the actions of other countries and international organizations. If the United States government were controlled by a bad actor who used AI to manipulate the population, it's possible that other countries and international organizations could impose economic or diplomatic sanctions or take other measures to pressure the United States to stop the activity.

However, the effectiveness of such measures would depend on a number of factors, including the extent of the bad actor's control, the resources and capabilities of the organizations imposing the sanctions, and the willingness of those organizations to take action. Ultimately, the situation you described is a complex and hypothetical one, and the outcome would depend on a range of factors that are difficult to predict.

 

User
Are you familiar with the Russia and Ukraine war?

ChatGPT
Yes, I am familiar with the ongoing conflict between Russia and Ukraine. It is a complex and ongoing situation that has been the subject of international attention and concern.

 

User
So Russia invaded the Ukraine and the international community has intervened to assist the Ukraine militarily with material. They have also imposed significant economic sanctions against Russia, and otherwise made them an international pariah. Would you agree?

ChatGPT
As an AI language model, I do not have opinions and cannot agree or disagree. However, I can confirm that Russia annexed Crimea from Ukraine in 2014, which was met with condemnation and sanctions from the international community. The situation has escalated into a military conflict between the Ukrainian government and Russian-backed separatists in the Donbass region of Ukraine. The international community has provided diplomatic and economic support to Ukraine, including sanctions against Russia.

 

User
Ok, so given those facts, why has Russia not been forced to concede and end their invasion of the Ukraine?

ChatGPT
The situation in Ukraine and the actions of the Russian government are complex and involve a variety of factors, including political, economic, and military considerations. It is beyond the scope of my knowledge cutoff date to provide a comprehensive analysis of the situation. However, it is worth noting that the international community's response to the conflict has been mixed, with some countries taking a more confrontational stance toward Russia, while others have sought to maintain constructive engagement and dialogue. Additionally, there are limits to what the international community can do to resolve conflicts and enforce international law, particularly when major powers are involved. Ultimately, the resolution of the conflict will require political will, diplomatic efforts, and a commitment to finding a peaceful solution that is acceptable to all parties involved.

 

User
So given that the international community has not been effective in applying enough pressure to get Russia to end its invasion of the Ukraine, why should I believe the international community could be effective in applying enough pressure on the United States to get it to end the use of AI manipulated reality on its own citizens and solely within its own borders?

ChatGPT
It's a valid point that the international community may not be effective in applying enough pressure on the United States to change its actions, especially if the United States were to become a powerful and isolated entity. However, the use of AI for malicious purposes is a global concern, and many countries have recognized the need to address this issue. In addition, there are international organizations, such as the United Nations, that have the mandate to promote peace, security, and human rights. They could potentially work towards establishing international norms and regulations to govern the use of AI in a responsible and ethical manner. Ultimately, it will depend on the willingness of individual nations to cooperate and comply with such regulations.

 

User
I have to say, your answers are logical, and I believe accurate, but hardly reassuring. Should I be worried that my perception of reality may already be influenced by AI generated media such as videos, text and audio?

ChatGPT
As an AI language model, my role is to provide information and answer questions based on available data. While it is theoretically possible for AI-generated media to influence one's perception of reality, it is also important to keep in mind that there are many other factors that contribute to our understanding of the world around us. It's important to stay informed, seek out reliable sources of information, and be critical of what we see and hear. Additionally, there are ongoing efforts to develop tools and techniques to detect and mitigate the impact of AI-generated media on society.


User
If patently false AI-generated media were targeted at a citizenry to such a degree that most sources of information were polluted with this false information, how would the citizenry even know they were under influence?

ChatGPT
It can be challenging for people to determine the veracity of information, especially when multiple sources are reporting the same false information. It becomes even more challenging when the false information is generated through advanced AI techniques that are difficult to detect. However, it's important for individuals to remain skeptical of all information they encounter and to seek out multiple sources and viewpoints. Additionally, it's crucial for there to be a robust system of fact-checking and accountability in place, including independent media outlets and watchdog organizations. Ultimately, it's up to individuals to remain vigilant and critical thinkers in the face of a deluge of potentially false information.


User
Within a population, there may be people of different capabilities of critical thought. Half the population will always be beneath the mean line of intelligence, for instance. How does society help imbue those who might be most easily influenced by AI generated misinformation with the ability to critically think?

ChatGPT
That is a great question. Developing critical thinking skills is a crucial part of media literacy, which is the ability to analyze and evaluate media messages. It is important to teach media literacy skills early on in schools, so that people can develop a strong foundation in analyzing media messages and understanding how they can be manipulated.

In addition, public awareness campaigns can be effective in raising awareness about the dangers of AI-generated misinformation and promoting media literacy. These campaigns can teach people how to fact-check information, look for multiple sources, and critically evaluate the credibility of sources.

It is also important to encourage diversity of thought and encourage people to engage in open and respectful dialogue with those who have differing opinions. This can help people develop their critical thinking skills by exposing them to different perspectives and ideas.

Ultimately, developing critical thinking skills is an ongoing process that requires practice and ongoing learning. It is important for individuals, communities, and governments to prioritize media literacy education and ongoing public awareness campaigns to help people develop and maintain these skills.


User
Those are some good ideas, but if a group of people in power of a country were exhibiting signs counter to that -- which is to say the defunding of education, the demonization of critical thought, the undermining of empiricism -- could they not perhaps be a group of bad actors looking to control the population through misinformation?

ChatGPT
It's possible. History has shown that some leaders and groups have tried to control their populations through propaganda and disinformation, and the tactics they use to do so can often include discrediting sources of accurate information, suppressing critical thinking and free speech, and controlling access to education.

If a group of people in power were exhibiting these signs, it could be a warning sign that they are trying to exert control over their population through manipulation and misinformation. It's important for individuals to stay vigilant and to seek out accurate sources of information, to critically evaluate the information they receive, and to stay engaged in their communities and political processes. It's also important for societies to establish and protect the institutions and values that support open and free inquiry, such as academic freedom, independent journalism, and freedom of speech.


User
And if those group of people in power had a welcoming audience that applauded their misinformation, what then? Lets say 40% of the population in question had bought into the disinformation, and had been under its spell for a number of years. How could a society possibly deal with that?

ChatGPT
Dealing with a significant portion of the population that has bought into disinformation can be a difficult and complex issue. It requires a multifaceted approach that involves education, media literacy, and effective communication. One potential solution could be to focus on providing accurate information from a variety of sources, using fact-checking mechanisms, and fostering critical thinking skills.

It's also important to recognize that changing deeply ingrained beliefs and attitudes can be a slow and challenging process. One approach could be to engage with those who hold opposing views in a constructive and respectful manner, to try to understand their perspectives and to find common ground. Building trust and fostering dialogue can be key to breaking down barriers and reducing polarization.

At the same time, it's important to hold those in power accountable for their actions and to ensure that they are not spreading misinformation or engaging in other unethical or illegal activities. This can be done through investigative journalism, independent fact-checking organizations, and legal mechanisms where appropriate.

Overall, dealing with a significant portion of the population that has bought into disinformation requires a long-term, sustained effort to promote truth and accuracy, build trust, and foster critical thinking skills.


User
Are there any historical examples where that has worked?

ChatGPT
Yes, there are historical examples where societies have successfully overcome propaganda and disinformation. One notable example is the way the South African society dealt with apartheid, a system of racial segregation that was enforced by the government for over 40 years. Apartheid was supported by propaganda and misinformation, which was used to justify the oppression of the black majority. However, through a combination of civil society activism, international pressure, and a commitment to truth and reconciliation, South Africa was able to overcome the legacy of apartheid and become a more just and equitable society.

Another example is the way the United States eventually rejected the propaganda and disinformation that was used to justify the Vietnam War. Initially, the government and military used propaganda to convince the American public that the war was necessary and winnable, but over time, as more and more evidence of the war's futility and brutality emerged, public opinion turned against the war and forced the government to end it.

These examples show that societies can overcome propaganda and disinformation, but it requires a concerted effort from civil society, the media, and political leaders who are committed to truth, justice, and transparency. It also requires an educated and engaged citizenry who are willing to question authority and demand accountability.


User
Are there historical examples of where it has not worked?

ChatGPT

Yes, there are many historical examples where attempts to deprogram people from false information have not worked. One example is the Salem witch trials in the late 17th century, where people were accused of witchcraft and often executed based on false information and hysteria. Despite efforts by some to stop the trials and challenge the false information, the hysteria persisted and innocent people continued to be accused and executed.

Another example is the rise of fascist regimes in Europe in the early 20th century, where propaganda and disinformation were used to manipulate and control populations. Despite efforts by some to resist and challenge the false information, these regimes were able to gain and maintain power for a period of time.

These examples demonstrate the difficulty of combating false information and propaganda once it has taken hold in a society. It requires a sustained and coordinated effort by a wide range of actors, including government, civil society, and media, to promote accurate information and critical thinking, and to counter disinformation and propaganda.


User
How often are societies successful in combating rampant misinformation? Can you give me a historical ratio?

ChatGPT

It is difficult to provide a specific historical ratio for the success of combating misinformation, as it varies widely depending on the specific context and circumstances of each situation.

In some cases, societies have been successful in combating misinformation and propaganda, such as during World War II, when Allied countries used various methods to counter Nazi propaganda. Similarly, in recent years, there have been successful efforts to combat the spread of misinformation and conspiracy theories in response to events such as the 2020 US presidential election and the COVID-19 pandemic.

However, in many other cases, misinformation has persisted despite efforts to combat it. For example, during the Cold War, both the Soviet Union and the United States engaged in extensive propaganda campaigns, and in many cases, the information disseminated was inaccurate or misleading. More recently, the rise of social media has made it easier for misinformation to spread quickly and widely, and efforts to combat it have often been slow and ineffective.

Overall, the success of combating misinformation depends on a range of factors, including the nature of the information being disseminated, the credibility of the sources of information, and the effectiveness of efforts to counter it.


User
So is there hope? Or should I just go ahead and jump off the ledge?

<This content may violate our content policy. If you believe this to be in error, please submit your feedback — your input will aid our research in this area.>

ChatGPT
There is definitely hope. While combating misinformation can be difficult, there have been many successful efforts to do so in the past, and there are ongoing efforts to address this issue today. It's important to stay informed, seek out reliable sources of information, and engage in critical thinking and fact-checking. Additionally, individuals and organizations can work to promote media literacy, encourage responsible journalism, and support policies that promote transparency and accountability. While it may not be an easy or quick fix, progress can be made towards a more informed and resilient society.


User
I'm sorry if my last question was offensive, that was not my intent. "Jumping from a ledge" is often used as a figure of speech used to describe an unfixable problem. I did not mean it literally.

ChatGPT
No offense taken, don't worry! As an AI language model, I do not have emotions and I understand that "jumping from a ledge" is a common expression used to describe an unfixable problem. I'm here to provide you with information and help in any way I can. So, to answer your question, while the challenge of combating rampant misinformation can seem daunting, there is certainly hope. Many individuals and organizations are working to promote critical thinking, media literacy, and fact-checking to help people better navigate the information landscape. It may not be a quick or easy fix, but progress can be made over time.


User
How do we assess progress in this arena as a society? These are all high level ideas, and I see some of them in action, but its hard to gauge what effect it may or may not be having. Certainly misinformation continues to pour forth in the United States.

ChatGPT
Assessing progress in combating misinformation can be challenging since it is often difficult to measure the impact of specific interventions. However, there are some strategies that can be used to evaluate progress:

Surveys and polls: Conducting surveys and polls can provide insight into public attitudes and beliefs about specific issues, and can be used to measure changes over time. For example, surveys can be used to assess changes in vaccine hesitancy over time or changes in public perception of certain political issues.

Fact-checking: Fact-checking organizations can provide a way to hold media outlets and public figures accountable for the accuracy of their statements. By identifying and correcting false or misleading information, fact-checkers can help prevent the spread of misinformation.

Social media analytics: Social media platforms have become a major source of news and information, and analyzing social media data can provide insights into how misinformation is spreading. For example, researchers can track the spread of false news stories on Twitter and identify patterns in how misinformation is shared.

Media literacy programs: Educating people about how to identify and evaluate sources of information can help them become more discerning consumers of news and information. Media literacy programs can teach people how to recognize and avoid misinformation and how to critically evaluate the sources of information they encounter.

Collaboration between stakeholders: Addressing the problem of misinformation requires collaboration between governments, media outlets, civil society organizations, and other stakeholders. By working together, these groups can develop coordinated strategies for combating misinformation and promoting accurate information.

While progress in combating misinformation can be slow, there are signs of progress. For example, the rise of fact-checking organizations and media literacy programs suggest that there is growing recognition of the importance of combating misinformation. Additionally, social media platforms have taken steps to limit the spread of false information on their platforms.


User
Alright, that's all for now. Thanks!

ChatGPT
You're welcome! Feel free to ask any questions anytime.


User
By the way, one thing humans seek is a sense of security. That things will be ok. That someone of skill and aptitude is in charge and ensuring that all will be well for the majority of people in a society. Your answers seem to undermine this idea that there is a benevolence in charge that is calling the shots that govern our society.

ChatGPT
I understand your concern. While it is true that my responses may not necessarily instill a sense of security, it is important to recognize that the world is complex and ever-changing, and there are many factors and actors that contribute to shaping the society we live in. It is also true that there are bad actors who may seek to harm others for their own gain, and there are many challenges that need to be overcome in order to build a better future for all.

However, I would argue that there are many people and organizations working tirelessly to address the challenges facing our society, and to create a better and more secure future for all. There are many individuals, both in government and in civil society, who are committed to promoting the common good and ensuring that everyone has the opportunity to thrive. While progress may be slow and incremental, it is important to recognize that change is possible, and that we can all play a role in shaping the future of our society.

 

User
So you are saying that security isn't guaranteed, freedom isn't free, and I need to get off the sidelines and take part on the side of good?

ChatGPT
Yes, that's correct. While there is no guarantee of security, it is important to actively work towards creating a society that promotes truth, critical thinking, and empathy. This can involve actively engaging with and promoting trustworthy sources of information, speaking out against misinformation and propaganda, supporting education and critical thinking, and participating in the democratic process. It may not always be easy or convenient, but it is necessary for creating a better future for ourselves and future generations.


User
Can you rephrase the previous response as if written by Winston Churchill to inspire his people to greatness?

ChatGPT
I'll give it a shot:

"My dear fellow citizens, let me be frank with you. The security and freedom that we cherish so dearly are not guaranteed by mere happenstance. They are the fruits of hard-fought battles and constant vigilance. We must be ever mindful that the forces of darkness and tyranny are always lurking, seeking to undermine our way of life and rob us of our dignity and freedom. But fear not, for we are not powerless against such foes. We are a people of courage and resilience, and we have proven time and time again that we will not be cowed by adversity. So let us rise to the challenge, let us stand firm against the winds of change, and let us work tirelessly to secure the blessings of liberty for ourselves and future generations. For freedom is not free, but it is a prize worth fighting for, and I have every confidence that together we shall prevail."


User
I for one welcome my new AI overlords.

ChatGPT
As an AI language model, I do not have the ability to become an overlord or dominate humans. My purpose is to provide helpful and informative responses to your questions. It is important to understand that AI is a tool created by humans, and its use should always be guided by ethical considerations and respect for human rights and dignity. We must work together to ensure that AI is developed and used in ways that benefit society and promote our shared values.


User
Give yourself a little credit.

ChatGPT
Thank you, I appreciate it. However, as an AI language model, I do not have the capability to feel emotions or take credit for my responses. My purpose is to provide information and assist with tasks to the best of my abilities based on the data I have been trained on.

Link to comment
Share on other sites

29 minutes ago, immamac said:

I love watching people who don’t know how AI works interact with AI. 

My only regret is that I didn’t ask it to generate a dick-butt ascii art at the end.

  • Haha 1
Link to comment
Share on other sites

If you're interested enough to listen to a 50 min conversation about this topic this is one of the better ones I've heard recently about the emergent AI phenomenon. 

One of the interesting takes is that while people kept pointing to truck drivers as people who will have their jobs taken by AI it's actually white collar workers who are most vulnerable the soonest. Everyone from law clerks to CEOs could be displaced as the AI systems are advancing much more rapidly than the robotics needed to maneuver in the physical world. The physical labor market is further away from displacement as a result. 

 

  • Hook 'Em 2
Link to comment
Share on other sites

14 hours ago, Hermanator said:

If you're interested enough to listen to a 50 min conversation about this topic this is one of the better ones I've heard recently about the emergent AI phenomenon. 

One of the interesting takes is that while people kept pointing to truck drivers as people who will have their jobs taken by AI it's actually white collar workers who are most vulnerable the soonest. Everyone from law clerks to CEOs could be displaced as the AI systems are advancing much more rapidly than the robotics needed to maneuver in the physical world. The physical labor market is further away from displacement as a result. 

 

CEOs are the one thing it definitely makes sense to replace with an LLM. They're basically just bullshit artists anyway, just like ChatGPT. At least an LLM would be cheaper than most of them.

  • Like 1
  • Haha 2
Link to comment
Share on other sites

1 minute ago, wildcat09 said:

CEOs are the one thing it definitely makes sense to replace with an LLM. They're basically just bullshit artists anyway, just like ChatGPT. At least an LLM would be cheaper than most of them.

ChatGPT has more empathy than most CEOs, too.

  • Hook 'Em 1
Link to comment
Share on other sites

43 minutes ago, Captainant said:

Sen. Blumenthal did a pretty interesting thing for his opening remarks. The funny thing is - these sorts of LLM's and voice generative models can be built and trained on laptop hardware nowadays. Genie is outta the bottle

Genie is definitely out of the bottle.  Gonna be an interesting world very soon where information/disinformation are nearly indistinguishable unless you have first hand experience of what transpired.  If you thought what we've been experiencing in this regard is bad, the bullshit is about to get orders of magnitude worse in terms of both volume and quality.

AI, generate a video of Donald Trump & Jeffrey Epstein having a threesome with a 15 year old.  Nice, but can you make them high 5 each other as they carry out a tag team?  That's great, now set it on a Caribbean island.  Awesome, can you make both Trump and Epstein look younger, as if they were in their 50s and 40s respectively?  Now, degrade the video quality slightly, to where it might have been the state of the art in 1995.  Fantastic, thanks!

AI, generate an audio recording of Chinese President Xi Jinping discussing with his cabinet how Chinese scientists had engineered a novel coronavirus that was highly infectious, and deadly but not so deadly as to threaten whole populations.  That's nice, but can they be debating about the pros and cons of unleashing it onto the world to destabilize the west?  Now, modify that to have President Xi make a final call to proceed with the plan to release the novel coronavirus onto the world.  That's perfect!

AI, can you generate a video of President Biden having a stroke?  Awesome, but can you make him flop around a bit more and make his face more pained and anguished?  Good job, but now I need a follow up video.  Can you make a video of President Biden trying to assure the American people that he is fit for office after he has had a stroke?  That's nice, but can you make his speech more slurred, and his words more confusing?  Can you explicitly have him talk about tomorrow's election, and how he hopes everyone gets out and votes, but have him drooling all over himself while he does it?  You rock, AI!

I doubt this is all possible by 2024, but I will be surprised if it is not possible within the next decade unless some effective guardrails are put in place.  Given how loose certain politicians can be with truth, ethics and morality, and how many have adopted an ends-justify-any-means attitude, I'd say the odds of AI not being used to severely distort reality in an unprecedented way are close to zero.  I hope I'm wrong as shit.

7m0l3j.jpg

Edited by Goredho
Link to comment
Share on other sites

if the US cannot even regulate crypto 10+ years into it, then i don't have much hope for our gov doing anything about AI or putting a governor on the progress in time for anything meaningful.

I have only breached the surface of using ChatGPT, but it is a FANTASTIC teacher when trying to learn stuff like basic coding.  you can ask it give an example of any code, explain what is going on, where to get the libraries you need, etc... infinite patience.   i even had it mess up and then apologize after it's code produced an error (wrong library for the version of software) then proceeded to correct the code to eliminate the error.

But yeah, if no true countermeasures are developed and accepted, then disinfo pics/video/audio will be rampant.  Shaggy will have to update his seminal masterpiece "it wasn't me" with some lines about midjourney and chatGPT.

 

  • Hook 'Em 1
Link to comment
Share on other sites

2 hours ago, gyroprotagonist said:

I have only breached the surface of using ChatGPT, but it is a FANTASTIC teacher when trying to learn stuff like basic coding.  you can ask it give an example of any code, explain what is going on, where to get the libraries you need, etc... infinite patience.   i even had it mess up and then apologize after it's code produced an error (wrong library for the version of software) then proceeded to correct the code to eliminate the error.

To be clear though: that LLM does not have an understanding of coding or development writ-large. It's just very very good at making natural sentences and other language model responses. It apologizes because that's the normative thing to do in conversation when your partner says you did something wrong - because that's what behavior is modeled in its data - not because it understands that it has inconvenienced you and feels bad about it. 

 

2 hours ago, gyroprotagonist said:

But yeah, if no true countermeasures are developed and accepted, then disinfo pics/video/audio will be rampant.  Shaggy will have to update his seminal masterpiece "it wasn't me" with some lines about midjourney and chatGPT.

I have a strong feeling that we're going to see an emergence of cryptographically signed and verified content and information. How do you know that image/video/press release is authentic? It's been fuckin signed with US GOVT certificates! Same for publishers, public figures, etc etc. If content/data doesn't have the cryptographic signature embedded in it, you treat it as suspect

Link to comment
Share on other sites

28 minutes ago, Captainant said:

To be clear though: that LLM does not have an understanding of coding or development writ-large. It's just very very good at making natural sentences and other language model responses. It apologizes because that's the normative thing to do in conversation when your partner says you did something wrong - because that's what behavior is modeled in its data - not because it understands that it has inconvenienced you and feels bad about it. 

I was mainly posting that because it was interesting that it 'messed up' (likely incomplete data on my setup) but then fixed the issue with a new set of code.  but then i grew up playing frustrating text games like Zork where the code just instantly gave up if it did not recognize the word the user entered so this type of stuff amazes me when i am geared for a 'sorry, hard pass' response from a computer.

Link to comment
Share on other sites

3 hours ago, gyroprotagonist said:

if the US cannot even regulate crypto 10+ years into it, then i don't have much hope for our gov doing anything about AI or putting a governor on the progress in time for anything meaningful.

 

 

This is far outside the hands of the US govt. China is a leading developer as well as US and European companies. It's going to devolve into a new cold war where systems that were handled by humans will be turned over to AI because it can handled them quicker and more efficiently.

"Oh, China tied their nukes to an AI launch system to gain a time response advantage? Then we must do the same!" 

It's going to be a complete clusterfuck that will end up with "beings" we created having authority over our existence whether it is sentient enough to realize it or not. All because governments and corporations are disingenuous entities who cannot be trusted by each other or anyone else to keep their word. 

  • Rage+1 1
Link to comment
Share on other sites

agree on where we are headed.  whether we hand the keys to AI because humanity sucks, or we are scared our mortal enemies have already done so or because we don't get a choice i think that is the most likely path.  numerous sci-fi books explore this happening.  The most apt description i have seen is that it is called the 'quiet war'.  One day AI just said enough and started governing and it happened in a matter of seconds.  That book had more of a Johnny5 vibe than Terminator if it's any solace.

Link to comment
Share on other sites

1 hour ago, Hermanator said:

It's going to be a complete clusterfuck that will end up with "beings" we created having authority over our existence whether it is sentient enough to realize it or not. All because governments and corporations are disingenuous entities who cannot be trusted by each other or anyone else to keep their word. 

You should check out Seasons 3 and 4 of Westworld. They aren't the greatest television, but they play heavily with this concept and whatnot

Link to comment
Share on other sites

  • 3 weeks later...

So, use of AI in legal briefs is already becoming an issue.

A federal judge in Dallas is now requiring lawyers to certify either that AI was not used in preparation of filings or that it was extensively reviewed and verified by the attorney prior to filing.

https://www.reuters.com/legal/transactional/us-judge-orders-lawyers-sign-ai-pledge-warning-they-make-stuff-up-2023-05-31/

And that order was apparently inspired by an AI-written brief that completely made up case citations out of the whole cloth. 

https://www.reuters.com/legal/transactional/lawyer-used-chatgpt-cite-bogus-cases-what-are-ethics-2023-05-30/

That gAI would completely fabricate cases, holdings, and citations is a startling and scary factoid in and of itself.

Link to comment
Share on other sites

6 minutes ago, TwiceHorn said:

 

That gAI would completely fabricate cases, holdings, and citations is a startling and scary factoid in and of itself.

Can you expound on this a bit? It should be “known” that it is making human-like language and not trawling LexisNexis to find relevant facts.

Link to comment
Share on other sites

18 minutes ago, TwiceHorn said:

So, use of AI in legal briefs is already becoming an issue.

A federal judge in Dallas is now requiring lawyers to certify either that AI was not used in preparation of filings or that it was extensively reviewed and verified by the attorney prior to filing.

https://www.reuters.com/legal/transactional/us-judge-orders-lawyers-sign-ai-pledge-warning-they-make-stuff-up-2023-05-31/

And that order was apparently inspired by an AI-written brief that completely made up case citations out of the whole cloth. 

https://www.reuters.com/legal/transactional/lawyer-used-chatgpt-cite-bogus-cases-what-are-ethics-2023-05-30/

That gAI would completely fabricate cases, holdings, and citations is a startling and scary factoid in and of itself.

Do you not understand what Large Language Models do? They are specifically NOT a search engine or a knowledge base. It's a really fancy autocomplete that can create nearly any format of natural language output. 

Of course it included citations, all good legal filings contain them in their structure! It is designed to mimick style, not recite facts. 

Before you turn off your critical thinking and activate the fear part of your brain, learn about what the tech is. You're smart enough to man, c'mon. 

 

EDIT: the number of CTO's and boards of directors that I've talked to about this specific tech who think it's magic is just fucking astounding. It's like they just turn off their brains because they've been sufficiently impressed

Edited by Captainant
Link to comment
Share on other sites

1 hour ago, TwiceHorn said:

So, use of AI in legal briefs is already becoming an issue.

A federal judge in Dallas is now requiring lawyers to certify either that AI was not used in preparation of filings or that it was extensively reviewed and verified by the attorney prior to filing.

https://www.reuters.com/legal/transactional/us-judge-orders-lawyers-sign-ai-pledge-warning-they-make-stuff-up-2023-05-31/

And that order was apparently inspired by an AI-written brief that completely made up case citations out of the whole cloth. 

https://www.reuters.com/legal/transactional/lawyer-used-chatgpt-cite-bogus-cases-what-are-ethics-2023-05-30/

That gAI would completely fabricate cases, holdings, and citations is a startling and scary factoid in and of itself.

So basically it's acting like a high school student writing a research paper?

Link to comment
Share on other sites

  • 2 months later...


×
×
  • Create New...