GPT-3: Its Nature, Scope, Limits, and Consequences

@article{Floridi2020GPT3IN,
  title={GPT-3: Its Nature, Scope, Limits, and Consequences},
  author={L. Floridi and Massimo Chiriatti},
  journal={Minds and Machines},
  year={2020},
  volume={30},
  pages={681-694}
}
In this commentary, we discuss the nature of reversible and irreversible questions, that is, questions that may enable one to identify the nature of the source of their answers. We then introduce GPT-3, a third-generation, autoregressive language model that uses deep learning to produce human-like texts, and use the previous distinction to analyse it. We expand the analysis to present three tests based on mathematical, semantic (that is, the Turing Test), and ethical questions and show that GPT… 

Playing Games with Ais: The Limits of GPT-3 and Similar Large Language Models

This article contributes to the debate around the abilities of large language models such as GPT-3, dealing with: firstly, evaluating how well GPT does in the Turing Test, secondly the limits of such

Intensional Artificial Intelligence: From Symbol Emergence to Explainable and Empathetic AI

TLDR
It is argued that an explainable artificial intelligence must possess a rationale for its decisions, be able to infer the purpose of observed behaviour, and be can to explain its decisions in the context of what its audience understands and intends, and a theory of meaning is proposed in which an agent should model the world a language describes rather than the language itself.

GPT-3 and InstructGPT: technological dystopianism, utopianism, and “Contextual” perspectives in AI ethics and industry

TLDR
Although OpenAI’s newest 2022 language model InstructGPT represents a small step in reducing toxic language and aligning GPT-3 with user intent, it does not provide any compelling solutions to manipulation or bias, and it is argued that solutions to address these issues must focus on organisational settings as a precondition for ethical decision-making in AI, and high-quality curated datasets as a precursor for less harmful language model outputs.

The great Transformer: Examining the role of large language models in the political economy of AI

TLDR
The role LLMs play in the political economy of AI as infrastructural components for AI research and development is explored, pointing out how they are intertwined with the business model of big tech companies and further shift power relations in their favour.

Compression, The Fermi Paradox and Artificial Super-Intelligence

The following briefly discusses possible difficulties in communication with and control of an AGI (artificial general intelligence), building upon an explanation of The Fermi Paradox and preceding

Ethics, Rules of Engagement, and AI: Neural Narrative Mapping Using Large Transformer Language Models

TLDR
A mechanism to harness the narrative output of large language models and produce diagrams or “maps” of the relationships that are latent in the weights of such models as the GPT-3 to provide insight into the organization of information, opinion, and belief in the model, which in turn provide means to understand intent and response in the context of physical distance.

Not Quite ‘Ask a Librarian’: AI on the Nature, Value, and Future of LIS

TLDR
This work asked the world's best language model, GPT‐3, fifteen difficult questions about the nature, value, and future of library and information science (LIS), topics that receive perennial attention from LIS scholars, providing an LIS‐tailored demonstration of the current performance of AI language models.

Self-recognition in conversational agents

TLDR
Methodology here constructs a textual version of the mirror test by placing the agent as the one and only judge to figure out whether the contacted one is an other, a mimicker, or oneself in an unsupervised manner, which is objective, self-contained, and devoid of humanness.

Do Artificial Intelligence Systems Understand?

TLDR
The conclusion states that it is not necessary to attribute understanding to a machine in order to explain its exhibited “intelligent” behavior; a merely syntactic and mechanistic approach to intelligence as a task-solving tool is needed to justify the range of operations that it can display in the current state of technological development.

Essential Features in a Theory of Context for Enabling Artificial General Intelligence

TLDR
The myriad pragmatic ways in which context has been used, or implicitly assumed, as a core concept in multiple AI sub-areas, such as representation learning and commonsense reasoning are synthesized.
...

References

SHOWING 1-10 OF 49 REFERENCES

The Radicalization Risks of GPT-3 and Advanced Neural Language Models

TLDR
GPT-3 demonstrates significant improvement over its predecessor, GPT-2, in generating extremist texts and its strength in generating text that accurately emulates interactive, informational, and influential content that could be utilized for radicalizing individuals into violent far-right extremist ideologies and behaviors.

How safe is our reliance on AI, and should we regulate it?

TLDR
Some of the writing that GPT-3 produced has been amazingly good; but some has also been racist, sexist, or otherwise perfidious text, due to the built-in biases of the data that it mines to teach itself to write.

What the Near Future of Artificial Intelligence Could Be

  • L. Floridi
  • Philosophy
    The 2019 Yearbook of the Digital Ethics Lab
  • 2020
In this article, I shall argue that AI’s likely developments and possible challenges are best understood if we interpret AI not as a marriage between some biological-like intelligence and engineered

Can GPT-3 Pass a Writer’s Turing Test?

TLDR
GPT-3 can internalize the rules of language without explicit programming or rules, and can sometimes fail at the simplest of linguistic tasks, but it can also excel at more difficult ones like imitating an author or waxing philosophical.

Common Sense, the Turing Test, and the Quest for Real AI

TLDR
Hector Levesque considers the role of language in learning, and identifies a possible mechanism behind common sense and the capacity to call on background knowledge: the ability to represent objects of thought symbolically.

The Winograd Schema Challenge

TLDR
This paper presents an alternative to the Turing Test that has some conceptual and practical advantages, and English-speaking adults will have no difficulty with it, and the subject is not required to engage in a conversation and fool an interrogator into believing she is dealing with a person.

Artificial Intelligence, Deepfakes and a Future of Ectypes

The art world is full of reproductions. Some are plain replicas, for example the Mona Lisa. Others are fakes or forgeries, like the BVermeers^ painted by Han van Meegeren that sold for $60 million

Computing Machinery and Intelligence

  • A. Turing
  • Philosophy
    The Philosophy of Artificial Intelligence
  • 1950
TLDR
The question, “Can machines think?” is considered, and the question is replaced by another, which is closely related to it and is expressed in relatively unambiguous words.

Digital’s Cleaving Power and Its Consequences

The digital is deeply transforming reality. This is much obvious and uncontroversial. The real questions are why, how, and so what. In each case, the answer is far from trivial and definitely open to

Introduction to Reversible Computing

TLDR
This book provides a compendium of both classical and recently developed results on reversible computing, exploring up-and-coming theories, techniques, and tools for the application of reversible computing the logical next step in the evolution of computing systems.