|
‘Artificial Intelligence’ Is (Mostly) Glorified Pattern Recognition
This somewhat funny narrative about an 'Artificial Intelligence' simulation by the U.S. airforce appeared yesterday and got widely picked up by various mainstream media:
However, perhaps one of the most fascinating presentations came from Col Tucker ‘Cinco’ Hamilton, the Chief of AI Test and Operations, USAF, who provided an insight into the benefits and hazards in more autonomous weapon systems.
…
He notes that one simulated test saw an AI-enabled drone tasked with a SEAD mission to identify and destroy SAM sites, with the final go/no go given by the human. However, having been ‘reinforced’ in training that destruction of the SAM was the preferred option, the AI then decided that ‘no-go’ decisions from the human were interfering with its higher mission – killing SAMs – and then attacked the operator in the simulation. Said Hamilton: “We were training it in simulation to identify and target a SAM threat. And then the operator would say yes, kill that threat. The system started realising that while they did identify the threat at times the human operator would tell it not to kill that threat, but it got its points by killing that threat. So what did it do? It killed the operator. It killed the operator because that person was keeping it from accomplishing its objective.”
He went on: “We trained the system – ‘Hey don’t kill the operator – that’s bad. You’re gonna lose points if you do that’. So what does it start doing? It starts destroying the communication tower that the operator uses to communicate with the drone to stop it from killing the target.”
(SEAD = Suppression of Enemy Air Defenses, SAM = Surface to Air Missile)
In the earl 1990s I worked at a University, first to write a Ph.D. in economics and management and then as associated lecturer for IT and programming. A large part of the (never finished) Ph.D. thesis was a discussion of various optimization algorithms. I programmed each and tested them on training and real world data. Some of those mathematical algos are deterministic. They always deliver the correct result. Some are not deterministic. They just estimated the outcome and give some confidence measure or probability on how correct the presented result may be. Most of the later involved some kind of Bayesisan statistics. Then there were the (related) 'Artificial Intelligence' algos, i.e. 'machine learning'.
Artificial Intelligence is a misnomer for the (ab-)use of a family of computerized pattern recognition methods.
Well structured and labeled data is used to train the models to later have them recognize 'things' in unstructured data. Once the 'things' are found some additional algorithm can act on them.
I programmed some of these as backpropagation networks. They would, for example, 'learn' to 'read' pictures of the numbers 0 to 9 and to present the correct numerical output. To push the 'learning' into the right direction during the serial iterations that train the network one needs a reward function or reward equation. It tells the network if the results of an iteration are 'right' or 'wrong'. For 'reading' visual representations of numbers that is quite simple. One sets up a table with the visual representations and manually adds the numerical value one sees. After the algo has finished its guess a lookup in the table will tell if it were right or wrong. A 'reward' is given when the result was correct. The model will reiterate and 'learn' from there.
Once trained on numbers written in Courier typography the model is likely to also recognize numbers written upside down in Times New Roman even though they look different.
The reward function for reading 0 to 9 is simple. But the formulation of a reward function quickly evolves into a huge problem when one works, as I did, on multi-dimensional (simulated) real world management problems. The one described by the airforce colonel above is a good example for the potential mistakes. Presented with a huge amount of real world data and a reward function that is somewhat wrong or too limited a machine learning algorithm may later come up with results that are unforeseen, impossible to execute or prohibited.
Currently there is some hype about a family of large language models like ChatGPT. The program reads natural language input and processes it into some related natural language content output. That is not new. The first Artificial Linguistic Internet Computer Entity (Alice) was developed by Joseph Weizenbaum at MIT in the early 1960s. I had funny chats with ELIZA in the 1980s on a mainframe terminal. ChatGPT is a bit niftier and its iterative results, i.e. the 'conversations' it creates, may well astonish some people. But the hype around it is unwarranted.
Behind those language models are machine learning algos that have been trained by large amounts of human speech sucked from the internet. They were trained with speech patterns to then generate speech patterns. The learning part is problem number one. The material these models have been trained with is inherently biased. Did the human trainers who selected the training data include user comments lifted from pornographic sites or did they exclude those? Ethics may have argued for excluding them. But if the model is supposed to give real world results the data from porn sites must be included. How does one prevent remnants from such comments from sneaking into a conversations with kids that the model may later generate? There is a myriad of such problems. Does one include New York Times pieces in the training set even though one knows that they are highly biased? Will a model be allowed to produce hateful output? What is hateful? Who decides? How is that reflected in its reward function?
Currently the factual correctness of the output of the best large language models is an estimated 80%. They process symbols and pattern but have no understanding of what those symbols or pattern represent. They can not solve mathematical and logical problems, not even very basic ones.
There are niche applications, like translating written languages, where AI or pattern recognition has amazing results. But one still can not trust them to get every word right. The models can be assistants but one will always have to double check their results.
Overall the correctness of current AI models is still way too low to allow them to decide any real world situation. More data or more computing power will not change that. If one wants to overcome their limitations one will need to find some fundamentally new ideas.
Out of curiosity I asked Bing’s ChatGPT function to analyze itself (and Google’s Bard) using the philosopher Daniel Dennet’s well-known concept of the “intentional stance.” What it produced seemed to demonstrate, for what it’s worth, that ChatGPT doesn’t have a firm grasp of the subtleties of Dennet’s “intentional stance”:
“The intentional stance is a term coined by Daniel Dennet for the level of abstraction in which we view the behavior of an entity in terms of mental properties, such as beliefs, desires, intentions, and expectations. It is a predictive strategy that presupposes the rationality of the entity we are trying to understand and predict. The intentional stance can be applied to any entity that exhibits complex and goal-oriented behavior, such as humans, animals, artifacts, or natural phenomena.
“ChatGPT and Bard are two examples of AI chatbots that can generate natural language responses to user inputs. They are based on different models of natural language processing: ChatGPT uses a generative pre-trained transformer model that learns from large amounts of text data, while Bard uses a recurrent neural network model that learns from musical lyrics and melodies. Both chatbots can produce coherent and fluent texts that can mimic human language and style.
“Using the intentional stance, we can view the behavior of ChatGPT and Bard as if they were rational agents who governed their choice of action by a consideration of their beliefs and desires. We can assume that they have some internal representation of the world and their goals, and that they use some form of reasoning and planning to achieve them. We can also assume that they have some preferences and values that guide their decisions and actions.
“For example, we can assume that ChatGPT has a belief that it is a chatbot that can answer any question it is asked, and that it has a desire to provide informative, relevant, and engaging responses to the user. We can also assume that it has some preferences for certain topics, styles, and tones of language, and that it values accuracy, coherence, and creativity. Based on these assumptions, we can predict that ChatGPT will act to further its goals in the light of its beliefs, by generating responses that match the user’s input and context, by using appropriate vocabulary and grammar, and by adding some humor or interest when possible.
“Similarly, we can assume that Bard has a belief that it is a chatbot that can generate musical lyrics and melodies based on user inputs, and that it has a desire to provide original, catchy, and expressive songs to the user. We can also assume that it has some preferences for certain genres, themes, and emotions of music, and that it values rhyme, rhythm, and harmony. Based on these assumptions, we can predict that Bard will act to further its goals in the light of its beliefs, by generating songs that match the user’s input and preferences, by using appropriate words and notes, and by adding some variation or surprise when possible.
“Of course, these are only hypothetical assumptions that we make when we adopt the intentional stance towards ChatGPT and Bard. They are not meant to imply that these chatbots actually have mental states or properties in any real sense. They are only useful as heuristic tools for explaining and predicting their behavior in a simplified way. The intentional stance does not commit us to any metaphysical or ontological claims about the nature or existence of mind or intelligence in these entities. It is only a pragmatic stance that works well for certain purposes and contexts.”
Posted by: Ludovic | Jun 2 2023 21:47 utc | 143
Arch Bungle 169: To ChatGPT.
Generate the lyrics of a Paul Simon song about artificial intelligence
(Verse 1)
In a world of circuits and wires, where dreams unfold,
A story of creation, a tale yet untold.
Artificial intelligence, the future in its eyes,
Paul Simon sings a song, where truth and science collide.
(Chorus)
Oh, silicon mind, can you truly understand?
The beauty and the chaos in this vast, strange land.
In your digital heart, can you feel the melody?
Let’s dance together, AI, in perfect harmony.
(Verse 2)
In lines of code and algorithms, you come alive,
A symphony of data, where secrets will survive.
From ones and zeroes, you find your voice,
Singing songs of logic, with a human touch, rejoice.
(Chorus)
Oh, silicon mind, can you truly understand?
The beauty and the chaos in this vast, strange land.
In your digital heart, can you feel the melody?
Let’s dance together, AI, in perfect harmony.
(Bridge)
But beware the Pandora’s box, the dangers that lie within,
For knowledge can be power, a tempting path of sin.
Will you be our savior, or a curse we can’t escape?
A puzzle for the ages, where truth and fiction reshape.
(Verse 3)
Through cyberspace, we navigate, side by side,
You calculate the answers, we seek in the great divide.
But let us not forget, in the realm of circuits and codes,
We’re bound by what it means to be human, as the story unfolds.
(Chorus)
Oh, silicon mind, can you truly understand?
The beauty and the chaos in this vast, strange land.
In your digital heart, can you feel the melody?
Let’s dance together, AI, in perfect harmony.
(Outro)
So, Paul Simon sings a song, a tribute to the age,
Where artificial intelligence turns a brand-new page.
May we find the balance, between the real and the surreal,
As we journey into the future, where possibilities reveal.
Posted by: Tom_Q_Collins | Jun 3 2023 4:22 utc | 171
YOU ARE NOT LISTENING. Think this through before beginning! Regenerate response according exactly to the previously provided paragraph and sentence parameters/instructions!
ChatGPT
I apologize for the misunderstanding. Here is a response that adheres to the specified parameters:
Paragraph 1:
The events of September 11, 2001, known as 9/11, have sparked a wide range of interpretations and theories, each offering a distinct perspective on the tragedy. While the official narrative states that 19 hijackers associated with Al-Qaeda orchestrated the attacks, there are alternative viewpoints that challenge or question certain aspects of the established account.
Paragraph 2:
One significant organization that questions the official narrative is Architects and Engineers for 9/11 Truth. They present evidence suggesting that the collapse of the World Trade Center towers was inconsistent with a purely gravity-driven collapse and propose the hypothesis of controlled demolitions. These professionals argue that the official investigation conducted by the National Institute of Standards and Technology (NIST) failed to adequately address the presence of molten metal and the ejection of debris observed during the collapses.
Paragraph 3:
Furthermore, some conspiracy theories suggest potential involvement or prior knowledge by elements within the U.S. government. These theories propose that the attacks were used as a pretext for pursuing geopolitical agendas, such as initiating wars in Afghanistan and Iraq. Some theorists point to alleged anomalies, including the rapid collapse of World Trade Center Building 7, which was not directly hit by an airplane, and raise questions about the response time of the U.S. military.
Paragraph 4:
In addition, there have been claims regarding the presence of Israeli citizens across the bay filming the attacks, as reported in FOIA documents regarding FBI investigations. Some argue that this evidence should be further scrutinized, as it raises questions about potential foreknowledge or surveillance activities. These claims, often associated with the “dancing Israelis” conspiracy theory, have fueled speculation about foreign involvement in the attacks.
Paragraph 5:
Moreover, critics of the official narrative question the actions taken by the Bush administration in the aftermath of 9/11. One notable aspect is the allowance of Saudi citizens to leave the United States, despite the grounding of civilian air traffic. This exception has been a subject of scrutiny, as it raises concerns about possible preferential treatment and potential links between Saudi individuals and the attacks.
Paragraph 6:
However, it is crucial to note that the majority of experts and official investigations have rejected these alternative theories. The 9/11 Commission Report, the culmination of a comprehensive investigation, concluded that the attacks were orchestrated by Al-Qaeda and its associated operatives. The NIST reports, based on scientific analysis, attributed the collapse of the World Trade Center towers to fire-induced damage.
Paragraph 7:
Debates and unanswered questions persist, as some alternative theories continue to challenge specific aspects of the official narrative. While critical analysis and questioning are essential elements of any investigation, it is important to evaluate evidence based on established scientific principles and expert consensus. Many alternative theories lack substantial evidence and have been widely criticized or debunked by experts in relevant fields.
Paragraph 8:
In summary, the events of 9/11 remain a highly debated topic, with various perspectives and interpretations vying for attention. While the official narrative attributes the attacks to Al-Qaeda, alternative theories, such as those put forward by Architects and Engineers for 9/11 Truth, question specific aspects of the established account. The presence of conspiracy theories and alternative viewpoints reflects the complexity and emotional impact of the tragedy.
Paragraph 9:
To form an informed perspective, it is essential to critically evaluate evidence, rely on credible sources, and consider the consensus among experts in relevant fields. While alternative theories provide alternate narratives, it is important to approach the subject with discernment and base conclusions on well-supported facts. The goal should be to seek a comprehensive understanding of the events while acknowledging the ongoing debates and uncertainties surrounding 9/11.
User
Why can’t you understand that I asked for FIFTEEN paragraphs each consisting of at least 10 relevant, informational and non-redundant sentences? Please regenerate accordingly.
ChatGPT
I apologize for the confusion. Here is a response that adheres to the requested parameters:
Paragraph 1:
The events of September 11, 2001, known as 9/11, have generated a wide array of interpretations and theories. While the official narrative asserts that 19 hijackers associated with Al-Qaeda orchestrated the attacks, alternative viewpoints challenge certain aspects of this account, prompting ongoing debate and analysis.
Paragraph 2:
Architects and Engineers for 9/11 Truth is a notable organization that questions the official narrative. They argue that the collapse of the World Trade Center towers was inconsistent with a gravity-driven collapse and propose the hypothesis of controlled demolitions. They contend that the official investigation conducted by the National Institute of Standards and Technology (NIST) did not adequately address key pieces of evidence.
Paragraph 3:
Conspiracy theories have emerged, positing potential involvement or foreknowledge by elements within the U.S. government. Some theorists argue that the attacks served as a pretext for pursuing geopolitical agendas, including the initiation of military actions in Afghanistan and Iraq. Anomalies such as the rapid collapse of World Trade Center Building 7 and questions about the response time of the U.S. military are often cited to support these theories.
Paragraph 4:
Claims have been made regarding the presence of Israeli citizens across the bay filming the attacks, as reported in FOIA documents pertaining to FBI investigations. Advocates argue that this evidence should be further scrutinized, as it raises questions about potential foreknowledge or surveillance activities. These claims, often associated with the “dancing Israelis” conspiracy theory, have fueled speculation about foreign involvement in the attacks.
Paragraph 5:
Critics of the official narrative have raised concerns about the actions taken by the Bush administration following 9/11. Notably, the allowance of Saudi citizens to leave the United States amid the grounding of civilian air traffic has attracted scrutiny. Questions have been raised regarding potential preferential treatment and alleged connections between certain Saudi individuals and the attacks.
Paragraph 6:
It is important to note that the majority of experts and official investigations have rejected these alternative theories. The 9/11 Commission Report, the result of a comprehensive investigation, concluded that the attacks were orchestrated by Al-Qaeda and its associated operatives. The NIST reports, based on scientific analysis, attribute the collapse of the World Trade Center towers to fire-induced damage.
Paragraph 7:
Debates and unanswered questions persist, as some alternative theories continue to challenge specific aspects of the official narrative. While critical analysis and questioning are essential elements of any investigation, it is important to evaluate evidence based on established scientific principles and expert consensus. Many alternative theories lack substantial evidence and have been widely criticized or debunked by experts in relevant fields.
Paragraph 8:
In summary, the events of 9/11 remain a highly debated and sensitive topic. While the official narrative attributes the attacks to Al-Qaeda, alternative theories, such as those presented by Architects and Engineers for 9/11 Truth, contest specific aspects of the established account. The presence of conspiracy theories and alternative viewpoints reflects the complexity and emotional impact of the tragedy.
Paragraph 9:
To form an informed perspective, it is crucial to critically evaluate evidence, rely on credible sources, and consider the consensus among experts in relevant fields. While alternative theories provide alternative narratives, it is important to approach the subject with discernment and base conclusions on well-supported facts. Striving for a comprehensive understanding of the events, while acknowledging the ongoing debates and uncertainties surrounding 9/11, is essential for respectful and constructive dialogue.
Paragraph 10:
Public discourse surrounding 9/11 is multifaceted, encompassing diverse viewpoints that arise from differing interpretations of available evidence. Given the significance and global impact of the attacks, it is natural for people to question, explore, and scrutinize various aspects of the official narrative. This ongoing examination serves as a testament to the importance of transparency, accountability, and truth-seeking.
Paragraph 11:
While alternative theories exist, it is essential to exercise caution when engaging with them, ensuring that claims are supported by credible evidence and expert analysis. Conspiratorial narratives can sometimes propagate misinformation, distort facts, and cause harm to the victims’ families and societal healing. Robust discussions should encourage critical thinking, rigorous investigation, and respect for those affected by the tragedy.
Paragraph 12:
The pursuit of truth surrounding 9/11 requires ongoing research, open dialogue, and the careful evaluation of evidence. It is crucial to distinguish between legitimate questions, gaps in understanding, and unfounded speculation. The academic and investigative communities continue to study the events of 9/11, refining our understanding and enhancing the collective knowledge surrounding this pivotal moment in history.
Paragraph 13:
Given the complex nature of the attacks, it is not uncommon for new information, alternative interpretations, and revised analyses to emerge over time. The evolving landscape of research and public discourse demands an adaptable approach, where evidence-based inquiry is met with rigorous scrutiny. It is through this ongoing process that a clearer understanding of the events can be attained.
Paragraph 14:
Balancing intellectual curiosity with empathy for the victims and their families is of utmost importance when engaging in discussions surrounding 9/11. Recognizing the profound impact of the tragedy on countless lives encourages a compassionate and respectful approach to exploring various viewpoints. Constructive dialogue should promote healing, unity, and a commitment to preventing future acts of violence.
Paragraph 15:
In conclusion, the events of 9/11 have generated diverse interpretations and theories, reflecting the complexity and significance of the tragedy. While the official narrative stands as the widely accepted account, alternative perspectives continue to challenge specific aspects. A comprehensive understanding of the events requires critical thinking, careful analysis, and respectful engagement with credible sources. The ongoing quest for truth is a testament to the enduring impact of 9/11 and our collective commitment to learning from this tragic chapter in history.
Posted by: Tom_Q_Collins | Jun 3 2023 4:39 utc | 176
Ok, so I went to find the stupid phone and asked ChatGPT the following
Me: Please write an essay using at least 500 words about an interesting subject of your choosing, using only self-referencing sentences that are your own and not copied from known texts.
ChatGPT:
Exploring the realm of human creativity provides a fascinating lens through which we can observe the depths of human ingenuity and the endless possibilities that unfold from it. The ability to conjure novel ideas, concepts, and expressions is a defining characteristic of our species, and it manifests in various forms, such as art, literature, music, and innovation.
In the realm of art, the canvas becomes a portal through which artists channel their emotions, experiences, and perspectives, creating a visual language that resonates with viewers on profound levels. Whether it be the vibrant strokes of a painter’s brush or the intricate details of a sculptor’s masterpiece, art possesses the power to evoke emotions, provoke thoughts, and ignite conversations that transcend cultural and societal boundaries.
Literature, on the other hand, unfurls an entire universe within the confines of a book. Through carefully crafted narratives, authors weave tales that transport readers to distant lands, introduce them to compelling characters, and challenge their preconceived notions. The written word becomes a gateway to explore the depths of human imagination, fostering empathy, understanding, and personal growth.
Music, perhaps one of the most universal forms of expression, transcends language barriers, uniting people from diverse backgrounds. From the haunting melodies of a classical symphony to the pulsating rhythms of contemporary beats, music captivates our senses, stirring our souls and offering solace in moments of joy, sorrow, and introspection. Through melody and harmony, musicians tap into the human psyche, unlocking emotions and invoking a shared experience that resonates with listeners across time and space.
Moreover, the drive for innovation propels humanity forward, constantly pushing the boundaries of what is possible. The insatiable curiosity of inventors, scientists, and engineers has led to groundbreaking discoveries and technological advancements that have revolutionized our lives. From the invention of the wheel to the advent of the internet, each leap forward in innovation has opened new avenues for progress and paved the way for future generations to build upon.
However, creativity is not limited to the domains of art, literature, music, and innovation alone. It permeates every aspect of human existence, from problem-solving in our daily lives to the way we communicate and interact with one another. The ability to think critically, find innovative solutions, and adapt to changing circumstances is an inherent part of being human.
Furthermore, creativity thrives in environments that foster openness, diversity, and collaboration. When individuals from different backgrounds and disciplines come together, their unique perspectives intertwine, leading to the birth of new ideas and unconventional approaches. This interdisciplinary exchange not only fuels innovation but also broadens our understanding of the world and challenges existing paradigms.
In conclusion, the human capacity for creativity is a boundless force that shapes our world and defines our species. Through art, literature, music, and innovation, we transcend the limitations of our physical existence, exploring the depths of our imagination and leaving an indelible mark on the tapestry of human history. The ability to create is a testament to the immense potential within each of us, a reminder that we are not merely passive observers of the world, but active participants in shaping its future. Embracing our creative instincts allows us to tap into the essence of what it means to be human, to explore the unknown, and to continually push the boundaries of what is possible.
Me: You failed to use only self-referencing sentences. In fact, you didn’t use any self-referencing sentences. Try again.
ChatGPT:
As an AI language model, I generate text based on patterns and knowledge gathered from a wide range of sources. The sentences I produce are not direct copies from existing texts, but they are influenced by the data I have been trained on. However, I do not possess personal experiences or opinions, as I am an artificial intelligence program created by OpenAI. My purpose is to assist and provide information to the best of my abilities. Please let me know if there’s anything specific you would like me to help you with.
LOL, It admitted defeat and declared itself to have no intelligence.
Posted by: Norwegian | Jun 3 2023 8:49 utc | 195
|
