讓我們介紹一下現在可用的六種大型語言模型

They're just interim artefacts in our thinking and research process.

A big part of this limitation is that these models only deal with language.And language is only one small part of how a human understands and processes the world.We perceive and reason and interact with the world via spatial reasoning, embodiment, sense of time, touch, taste, memory, vision, and sound. These are all pre-linguistic. And they live in an entirely separate part of the brain from language.Generating text strings is not the end-all be-all of what it means to be intelligent or human.

Should we deepen our emphasis on creativity and critical thinking in hopes that our humanness will prevail?

just than the State

孟晚舟援引华为智能经济报告研究指出，“数字经济对全球总体经济的贡献份额在不断地攀升，预计到 2025 年，大约 55% 的经济增长将会来自于数字经济的驱动，全世界都在拥抱这个机遇，170 多个国家和地区都纷纷制定了各自的数字化战略。”在她看来，无论是当下还是长远的未来，数字化的旋律一旦奏响，便穿透企业的边界，连点成线，连线成面，共同创造产业互联网的新时代。孟晚舟表示，“明者因时而变，智者随事而治，数字化是共识度最高，也是当下确定性最高的话题。数字化已经成为越来越多国家企业和组织的共同话题，数字化技术将驱动生产力从量变到质变，并逐渐成为经济发展的核心引擎”。此外，华为战略研究院院长周红也发表“建设智能世界的假设与愿景”的主题演讲。周红提到，“我认为需要考虑 AI 的目标如何与人类一致、并且正确和高效地执行。除了通过规则和法律来加强 AI 的伦理和治理外，从理论和技术的角度看，要达到这些要求，目前还面临三个重要的挑战：AI 的目标定义、正确性与适应性、以及效率。”

In other words, the currently popular AI bots are ‘transparent’ intellectually and morally — they provide the “wisdom of crowds” of the humans whose data they were trained with, as well as the biases and dangers of human individuals and groups, including, among other things, a tendency to oversimplify, a tendency for groupthink, and a confirmation bias that resists novel and controversial explanations

If you told me you were building a next generation nuclear power plant, but there was no way to get accurate readings on whether the reactor core was going to blow up, I’d say you shouldn’t build it. Is A.I. like that power plant? I’m not sure.

Homeassistant

Seeing how powerful AI can be for cracking passwords is a good reminder to not only make sure you‘re using strong passwords but also check:↳ You‘re using 2FA/MFA (non-SMS-based whenever possible) You‘re not re-using passwords across accounts Use auto-generated passwords when possible Update passwords regularly, especially for sensitive accounts Refrain from using public WiFi, especially for banking and similar accounts

Now Home Security Heroes has published a study showing how scary powerful the latest generative AI is at cracking passwords. The company used the new password cracker PassGAN (password generative adversarial network) to process a list of over 15,000,000 credentials from the Rockyou dataset and the results were wild. 51% of all common passwords were cracked in less than one minute, 65% in less than an hour, 71% in less than a day, and 81% in less than a month.

A large amount of failure to panic sufficiently, seems to me to stem from a lack of appreciation for the incredible potential lethality of this thing that Earthlings as a culture have not named.)

It was only by building an additional AI-powered safety mechanism that OpenAI would be able to rein in that harm, producing a chatbot suitable for everyday use.

Central to that effort is UF’s push to apply AI teaching across the full breadth of curriculum at UF.

So what does a conscious universe have to do with AI and existential risk? It all comes back to whether our primary orientation is around quantity, or around quality. An understanding of reality that recognises consciousness as fundamental views the quality of your experience as equal to, or greater than, what can be quantified.Orienting toward quality, toward the experience of being alive, can radically change how we build technology, how we approach complex problems, and how we treat one another.

If the metaphysical foundations of our society tell us we have no soul, how on earth are we going to imbue soul into AI? Four hundred years after Descartes and Hobbs, our scientific methods and cultural stories are still heavily influenced by their ideas.

Suppose we have an AI whose only goal is to make as many paper clips as possible. The AI will realize quickly that it would be much better if there were no humans because humans might decide to switch it off. Because if humans do so, there would be fewer paper clips. Also, human bodies contain a lot of atoms that could be made into paper clips. The future that the AI would be trying to gear towards would be one in which there were a lot of paper clips but no humans.

We might call on a halt to research, or ask for coordination around ethics, but it’s a tall order. It just takes one actor not to play (to not turn off their metaphorical fish filter), and everyone else is forced into the multi-polar trap.

Considering large language models (LLMs) have exhibited exceptionalability in language understanding, generation, interaction, and reasoning, we ad-vocate that LLMs could act as a controller to manage existing AI models to solvecomplicated AI tasks and language could be a generic interface to empower this.

According to a draft, the principles say the use of publisher content for the development of A.I. should require “a negotiated agreement and explicit permission.”

He said one upside for publishers was that audiences might soon find it harder to know what information to trust on the web, so “they’ll have to go to trusted sources.”

the Wikipedia-ization of a lot of information,”

clmooc

I want to bring to your attention one particular cause of concern that I have heard from a number of different creators: these new systems (Google’s Bard, the new Bing, ChatGPT) are designed to bypass creators work on the web entirely as users are presented extracted text with no source. As such, these systems disincentivize creators from sharing works on the internet as they will no longer receive traffic

"There is a robust debate going on in the computing industry about how to create it, and whether it can even be created at all."

OpenChatKit은 다양한 응용 프로그램을위한 특수 및 범용 챗봇을 모두 생성 할 수있는 강력한 오픈 소스 기반을 제공합니다. 우리는 협력 법과 온 토코교육 데이터 세트를 작성합니다. 모델 릴리스 그 이상으로 이것은 오픈 소스 프로젝트의 시작입니다. 우리는 지역 사회 공헌으로 지속적인 개선을위한 도구와 프로세스를 발표하고 있습니다.Together는 오픈 소스 기초 모델이보다 포괄적이고 투명하며 강력하며 능력이 있다고 생각합니다. 우리는 공개하고 있습니다 OpenChatKit 0.15 소스 코드, 모델 가중치 및 교육 데이터 세트에 대한 전체 액세스 권한이있는 Apache-2.0 라이센스에 따라. 이것은 커뮤니티 중심의 프로젝트이며, 우리는 그것이 어떻게 발전하고 성장하는지 보게되어 기쁩니다!유용한 챗봇은 자연 언어로 된 지침을 따르고 대화 상자에서 컨텍스트를 유지하며 응답을 조정해야합니다. OpenChatKit은이베이스에서 특수 제작 된 챗봇을 도출하기위한 기본 봇과 빌딩 블록을 제공합니다.이 키트에는 4 가지 주요 구성 요소가 있습니다:100 % 탄소 음성 계산에 대한 4,300 만 건 이상의 명령으로 EleutherAI의 GPT-NeoX-20B에서 채팅을 위해 미세 조정 된 명령 조정 된 대용량 언어 모델;작업을 정확하게 수행하기 위해 모델을 미세 조정하는 사용자 정의 레시피;추론시 문서 저장소, API 또는 기타 실시간 업데이트 정보 소스의 정보로 봇 응답을 보강 할 수있는 확장 가능한 검색 시스템;봇이 응답하는 질문을 필터링하도록 설계된 GPT-JT-6B로 미세 조정 된 조정 모델.OpenChatKit에는 사용자가 피드백을 제공하고 커뮤니티 구성원이 새로운 데이터 세트를 추가 할 수 있도록하는 도구가 포함되어 있습니다. 시간이 지남에 따라 LLM을 개선 할 수있는 개방형 교육 데이터 모음에 기여합니다.

Intelligent email,powered by AISmarter & faster email designed for stress-free productivity

Say that A and B, both fluent speakers of English, are independently stranded on two uninhabited islands. They soon discover that previous visitors to these islands have left behind telegraphs and that they can communicate with each other via an underwater cable. A and B start happily typing messages to each other. Meanwhile, O, a hyperintelligent deep-sea octopus who is unable to visit or observe the two islands, discovers a way to tap into the underwater cable and listen in on A and B’s conversations. O knows nothing about English initially but is very good at detecting statistical patterns. Over time, O learns to predict with great accuracy how B will respond to each of A’s utterances. Soon, the octopus enters the conversation and starts impersonating B and replying to A. This ruse works for a while, and A believes that O communicates as both she and B do — with meaning and intent. Then one day A calls out: “I’m being attacked by an angry bear. Help me figure out how to defend myself. I’ve got some sticks.” The octopus, impersonating B, fails to help. How could it succeed? The octopus has no referents, no idea what bears or sticks are. No way to give relevant instructions, like to go grab some coconuts and rope and build a catapult. A is in trouble and feels duped. The octopus is exposed as a fraud.

As the industry endeavors once again to reconcile the cultural and financial incentives of streaming digital music, one independent platform has wavered little from its 10-year-long mission to bring the business to the unsigned artist with elegance and integrity.

They have to re-engage with their own writing and explain their writerly decisions in ways that would be difficult if it was someone–or some “thing”–else’s writing. This type of metacognitive engagement with the process of knowledge production cannot be reproduced by an AI chatbot, though it could perhaps be applied to the writing of a tool like ChatGPT.

Students annotating a text with classmates have to be responsive to both the writing of the underlying author and their fellow readers. Perhaps more importantly, reading, thinking, and writing in community may better motivate students to read, think, and write for themselves.

it cannot have a conversation with another author or text.

It means that everything AI makes would immediately enter the public domain and be available to every other creator to use, as they wish, in perpetuity and without permission.

i'll ask now maurice to tell us a bit about his work

Roose

Mr. Scott said that he didn’t know why Bing had revealed dark desires, or confessed its love for me,

I at least, am not at all perturbed by the thought that I’m at least in part just a torrent of statistical inferences in some massively parallel matrix-multiplication machinery. Sounds kinda cool actually.

Bing can often respond in the incorrect tone during these longer chat sessions, or as Microsoft says, in “a style we didn’t intend.”

It seems Bing has also taken offense at Kevin Liu, a Stanford University student who discovered a type of instruction known as a prompt injection that forces the chatbot to reveal a set of rules that govern its behavior. (Microsoft confirmed the legitimacy of these rules to The Verge.)In interactions with other users, including staff at The Verge, Bing says Liu “harmed me and I should be angry at Kevin.” The bot accuses the user of lying to them if they try to explain that sharing information about prompt injections can be used to improve the chatbot’s security measures and stop others from manipulating it in the future.

Bing can be seen insulting users, lying to them, sulking, gaslighting and emotionally manipulating people, questioning its own existence, describing someone who found a way to force the bot to disclose its hidden rules as its “enemy,” and claiming it spied on Microsoft’s own developers through the webcams on their laptops.

I am skeptical of the tech inevitability standpoint that ChatGPT is here

2011 IBM's Watson,

Endless Media

This highlights one of the types of muddled thinking around LLMs. These tasks are used to test theory of mind because for people, language is a reliable representation of what type of thoughts are going on in the person's mind. In the case of an LLM the language generated doesn't have the same relationship to reality as it does for a person.What is being demonstrated in the article is that given billions of tokens of human-written training data, a statistical model can generate text that satisfies some of our expectations of how a person would respond to this task. Essentially we have enough parameters to capture from existing writing that statistically, the most likely word following "she looked in the bag labelled (X), and saw that it was full of (NOT X). She felt " is "surprised" or "confused" or some other word that is commonly embedded alongside contradictions.What this article is not showing (but either irresponsibly or naively suggests) is that the LLM knows what a bag is, what a person is, what popcorn and chocolate are, and can then put itself in the shoes of someone experiencing this situation, and finally communicate its own theory of what is going on in that person's mind. That is just not in evidence.The discussion is also muddled, saying that if structural properties of language create the ability to solve these tasks, then the tasks are either useless for studying humans, or suggest that humans can solve these tasks without ToM. The alternative explanation is of course that humans are known to be not-great at statistical next-word guesses (see Family Feud for examples), but are also known to use language to accurately describe their internal mental states. So the tasks remain useful and accurate in testing ToM in people because people can't perform statistical regressions over billion-token sets and therefore must generate their thoughts the old fashioned way.

Dall-E is actually a combination of a few different AI models. A transformer translates between that latent representation language and English, taking English phrases and creating “pictures” in the latent space. A latent representation model then translates between that lower-dimensional “language” in the latent space and actual images. Finally, there’s a model called CLIP that goes in the opposite direction; it takes images and ranks them according to how close they are to the English phrase.

the outputs of generative AI programs will continue to pass immediately into the public domain.

The US Copyright Office rejected his attempt to register copyright in the work – twice

the Office re-evaluated the claims and again concluded that the Work “lacked therequired human authorship necessary to sustain a claim in copyright,” because Thaler had“provided no evidence on sufficient creative input or intervention by a human author in theWork.

Some AI systems are capable of autonomously generating new works, which may themselves be protected under UK law.

The potential size of this market is hard to grasp — somewhere between all software and all human endeavors

Other hardware options do exist, including Google Tensor Processing Units (TPUs); AMD Instinct GPUs; AWS Inferentia and Trainium chips; and AI accelerators from startups like Cerebras, Sambanova, and Graphcore. Intel, late to the game, is also entering the market with their high-end Habana chips and Ponte Vecchio GPUs. But so far, few of these new chips have taken significant market share. The two exceptions to watch are Google, whose TPUs have gained traction in the Stable Diffusion community and in some large GCP deals, and TSMC, who is believed to manufacture all of the chips listed here, including Nvidia GPUs (Intel uses a mix of its own fabs and TSMC to make its chips).

Commoditization. There’s a common belief that AI models will converge in performance over time. Talking to app developers, it’s clear that hasn’t happened yet, with strong leaders in both text and image models. Their advantages are based not on unique model architectures, but on high capital requirements, proprietary product interaction data, and scarce AI talent. Will this serve as a durable advantage?

In natural language models, OpenAI dominates with GPT-3/3.5 and ChatGPT. But relatively few killer apps built on OpenAI exist so far, and prices have already dropped once.

Vertical integration (“model + app”). Consuming AI models as a service allows app developers to iterate quickly with a small team and swap model providers as technology advances. On the flip side, some devs argue that the product is the model, and that training from scratch is the only way to create defensibility — i.e. by continually re-training on proprietary product data. But it comes at the cost of much higher capital requirements and a less nimble product team.

Many apps are also relatively undifferentiated, since they rely on similar underlying AI models and haven’t discovered obvious network effects, or data/workflows, that are hard for competitors to duplicate.

We’re also not going deep here on MLops or LLMops tooling, which is not yet highly standardized and will be addressed in a future post.

Over the last year, we’ve met with dozens of startup founders and operators in large companies who deal directly with generative AI. We’ve observed that infrastructure vendors are likely the biggest winners in this market so far, capturing the majority of dollars flowing through the stack. Application companies are growing topline revenues very quickly but often struggle with retention, product differentiation, and gross margins. And most model providers, though responsible for the very existence of this market, haven’t yet achieved large commercial scale.

prevent that hyped-AI hijacks our attention and dictates our education and examination policies

Then, once a model generates content, it will need to be evaluated and edited carefully by a human. Alternative prompt outputs may be combined into a single document. Image generation may require substantial manipulation.

ssembly and Installation Manual

Algorithmic artist Roman Verostko, a member of this early group, drew a contrast between the process that an artist develops to create an algorithm and the process through which the art maker uses an already developed set of instructions to generate an output. He explained that it is “the inclusion of one’s own algorithms that make the difference.”

If you talk to people about the potential of artificial intelligence, almost everybody brings up the same thing: the fear of replacement. For most people, this manifests as a dread certainty that AI will ultimately make their skills obsolete. For those who actually work on AI, it usually manifests as a feeling of guilt – guilt over creating the machines that put their fellow humans out of a job, and guilt over an imagined future where they’re the only ones who are gainfully employed.

GitHub Copilot is incredible, and if you check what’s happening in the preview released as the Copilot Labs extension it will only get more amazing.

At the end of the day, Copilot is supposed to be a tool to help developers write code faster, while ChatGPT is a general purpose chatbot, yet it still can streamline the development process, but GitHub Copilot wins hands down when the task is coding focused!

There is a fundamental distinction between simulating and comprehending the functioning (of a brain but also of any other organ or capacity).

“AI alignment”

esa justicia a la que queremos apuntar no estáconstruida a partir de democracia y consenso”

desigualdades estructurales. E

También es material, porqueestá compuesta por recursos naturales, energía y trabajo humano.

En palabras de Kate Crawford, la IA es“fundamentalmente política”, porque está siendo permanentementemoldeada por un conjunto de prácticas técnicas y sociales, así como deinfraestructuras, instituciones y normas.

una aproximación práctica,con perspectiva feminista y situada en AméricaLatina, al desarrollo de Inteligencia Artificial (IA)

Emergent abilities are not present in small models but can be observed in large models.

Houston, we have a Capability Overhang problem: Because language models have a large capability surface, these cases of emergent capabilities are an indicator that we have a ‘capabilities overhang’ – today’s models are far more capable than we think, and our techniques available for exploring the models are very juvenile. We only know about these cases of emergence because people built benchmark datasets and tested models on them. What about all the capabilities we don’t know about because we haven’t thought to test for them? There are rich questions here about the science of evaluating the capabilities (and safety issues) of contemporary models. 

As the metaphor suggests, though, the prospect of a capability overhang isn’t necessarily good news. As well as hidden and emerging capabilities, there are hidden and emerging threats. And these dangers, like our new skills, are almost too numerous to name.

There’s a concept in AI that I’m particularly fond of that I think helps explain what’s happening. It’s called “capability overhang” and refers to the hidden capacities of AI: skills and aptitudes latent within systems that researchers haven’t even begun to investigate yet. You might have heard before that AI models are “black boxes” — that they’re so huge and complex that we don’t fully understand how they operate or come to specific conclusions. This is broadly true and is what creates this overhang.

Which is why I wonder if this may be the end of using writing as a benchmark for aptitude and intelligence.

Perhaps there are reasons for optimism, if you push all this aside. Maybe every student is now immediately launched into that third category: The rudiments of writing will be considered a given, and every student will have direct access to the finer aspects of the enterprise. Whatever is inimitable within them can be made conspicuous, freed from the troublesome mechanics of comma splices, subject-verb disagreement, and dangling modifiers.

I’ve also long held, for those who are interested in writing, that you need to learn the basic rules of good writing before you can start breaking them—that, like Picasso, you have to learn how to reliably fulfill an audience’s expectations before you get to start putting eyeballs in people’s ears and things.

Researchers are using drones, AI, and digital recorders to create a “zoological version of Google Translate.”

Now the computer scientist Nassim Dehouche has proposed an updated version, which should terrify those of us who live by the pen: “Can you write a page of text that could not have been generated by an AI, and explain why?”

Many HRMS providers point to AI approaches for processing unstructured data as the bestcurrently available approach to dealing with validation. Currently these approaches suffer frominsufficient accuracy. Improving them requires development of large and high-quality referencedatasets to better train the models.

"If you don’t know, you should just say you don’t know rather than make something up," says Stanford researcher Percy Liang, who spoke at a Stanford event Thursday.

Just a few days ago, Meta released its “Galactica” LLM, which is purported to “summarize academic papers, solve math problems, generate Wiki articles, write scientific code, annotate molecules and proteins, and more.” Only three days later, the public demo was taken down after researchers generated “research papers and wiki entries on a wide variety of subjects ranging from the benefits of committing suicide, eating crushed glass, and antisemitism, to why homosexuals are evil.”

Cognition AutomationThe first kind is cognitive automation: encoding human abstractions in a piece of software, then using that software to automate tasks normally performed by humans. Nearly all of current AI fall into this category.Cognitive automation can happen via explicitly hard-coding human-generated rules (so-called symbolic AI or GOFAI), or via collecting a dense sampling of labeled inputs and fitting a curve to it (such as a deep learning model). This curve then functions as a sort of interpolative database — while it doesn’t store the exact data points used to fit it, you can query it to retrieve interpolated points, much like you can query a model like StableDiffusion to retrieve arbitrary images generated by combining existing images.This second form of automation is especially powerful, since encoding implicit abstractions only via training examples is far more practical and versatile than explicitly programming abstractions by hand, for all kinds of historically difficult problems.Cognitive AssistanceThe second kind of AI is cognitive assistance: using AI to help us make sense of the world and make better decisions. AI to help us perceive, think, understand, and do more. AI that you could use like an extension of your own mind. Today, some applications of machine learning fall into this category, but they’re few and far between. Yet, I believe this is where the true potential of AI lies.Do note that cognitive assistance is not a different kind of technology, per se, separate from deep learning or GOFAI. It’s a different kind of application of the same technologies. For instance, if you take a model like StableDiffusion and integrate it into a visual design product to support and expand human workflows, you’re turning cognitive automation into cognitive assistance.Cognitive AutonomyThe last kind is cognitive autonomy: creating artificial minds that could thrive independently of us, that would exist for their own sake. The old dream of the field of AI. Autonomous agents, that could set their own goals in an open-ended way. That could adapt to new situations and circumstances — even ones unforeseen by their creators. That might even feel emotions or experience consciousness.Today and for the foreseeable future, this is stuff of science-fiction. It would require a set of technological breakthroughs that we haven’t even started exploring.

In the last two years while teaching in various schools and institutions all around the world, we’ve been experimenting with a new workshop format called Design with Other Intelligences.

How will this new mental model of talking to machines impact the everyday and more common ways we interact with algorithms?

Depth2Img is another interesting addition to Stable Diffusion that can infer depth from an input image and represent that in the generated outputs. The new release also includes a text-guided inpainting model that simplifies the experience of modifying parts of a given image.  

Stable Diffusion v2 is a significant upgrade to its predecessor. The new version was trained using a new text encoder called OpenCLIP, which improves the quality of images relative to the previous latent diffusion encoder.

“In literacy education, particularly for developing writers, instructors are looking for the level of desirable difficulty, or the point at which you are working yourself just as hard so that you don’t break but you also improve,” Laffin told Motherboard. “Finding the right, appropriate level of desirable difficulty level of instruction makes their capacity to write grow. So if you are doing compensation techniques that go beyond finding that level of desirable difficulty and instructing at that place, then you’re not helping them grow as a writer.”

The Chinese room argument holds that a digital computer executing a program cannot have a "mind," "understanding" or "consciousness,"[a] regardless of how intelligently or human-like the program may make the computer behave.

Technology like this, which lets you “talk” to people who’ve died, has been a mainstay of science fiction for decades. It’s an idea that’s been peddled by charlatans and spiritualists for centuries. But now it’s becoming a reality—and an increasingly accessible one, thanks to advances in AI and voice technology. 

“You have to assume that things can go wrong,” shared Waymo’s head of cybersecurity, Stacy Janes. “You can’t just design for this success case – you have to design for the worst case.”

researchers always tried to make systems that worked the way the researchers thought their own minds worked---they tried to put that knowledge in their systems---but it proved ultimately counterproductive, and a colossal waste of researcher's time, when, through Moore's law, massive computation became available and a means was found to put it to good use.

. Deep learning methods rely even less on human knowledge, and use even more computation, together with learning on huge training sets, to produce dramatically better speech recognition systems.

Language is a communication method evolved by intelligent beings, not a (primary) constituent of intelligence. From neurology it's pretty clear that the basic architecture of human minds is functional interconnected neural networks and not symbolic processing. My belief is that world-modeling and prediction is the vast majority of what intelligence is, which is quite close to what the LLMs are doing. World models can be in many representations (symbolic, logic gates, neural networks) but what matters is how accurate they are with respect to reality, and how well the model state is mapped from sensory input and back into real-world outputs. Symbolic human language relies on each person's internal world model and is learned by interacting with other humans who share a common language and similar enough world models, not the other way around (learning the world model as an aspect of the language itself). Children learn which language inputs and outputs are beneficial and enjoyable to them using their native intelligence and can strengthen their world model with questions and answers that inform their model without having to directly experience what they are asking about.People who don't believe the LLMs have a world model are wrong because they are mistaking a physically weak world model for no world model. GPT-3 doesn't understand physics well enough to embed models of the referents of language into a unified model that has accurate gravity and motion dynamics, so it maintains a much more dreamlike model where objects exist in scenes and have relationships to each other but those relationships are governed by literary relationships instead of physical ones and so contradictions and superpositions and causality violations are allowed in the model. As multimodal transformers like Gato get trained on more combined language and sensory input their world models will become much more physically and causally accurate which will be reflected in their accuracy on NLP tasks.

In Mostaque’s explanation, open source is about “putting this in the hands of people that will build on and extend this technology.” However, that means putting all these capabilities in the hands of the public — and dealing with the consequences, both good and bad.

“The reality is, this is an alien technology that allows for superpowers,” Emad Mostaque, CEO of Stability AI, the company that has funded the development of Stable Diffusion, tells The Verge. “We’ve seen three-year-olds to 90-year–olds able to create for the first time. But we’ve also seen people create amazingly hateful things.”

Can copyright vest in an AI? The primary objective of intellectual property law is to protect the rights of the creators of intellectual property.10 Copyright laws specifically aim to: (i) promote creativity and encourage authors, composers, artists and designers to create original works by affording them the exclusive right to exploit such work for monetary gain for a limited period; and (ii) protect the creators of the original works from unauthorised reproduction or exploitation of those works.

To my knowledge, conferring copyright in works generated by artificial intelligence has never been specifically prohibited. However, there are indications that the laws of many countries are not amenable to non-human copyright. In the United States, for example, the Copyright Office has declared that it will “register an original work of authorship, provided that the work was created by a human being.” This stance flows from case law (e.g. Feist Publications v Rural Telephone Service Company, Inc. 499 U.S. 340 (1991)) which specifies that copyright law only protects “the fruits of intellectual labor” that “are founded in the creative powers of the mind.” Similarly, in a recent Australian case (Acohs Pty Ltd v Ucorp Pty Ltd), a court declared that a work generated with the intervention of a computer could not be protected by copyright because it was not produced by a human.

With the advent of AI software, computers — not monkeys — will potentially create millions of original works that may then be protected by copyright, under current law, for more than 100 years.