They might want to add a new AI assistant developed by Google DeepMind to their arsenal. It can suggest tactics for soccer set-pieces that are even better than those created by professional club coaches.

The system, called TacticAI, works by analyzing a dataset of 7,176 corner kicks taken by players for Liverpool FC, one of the biggest soccer clubs in the world.

Corner kicks are awarded to an attacking team when the ball passes over the goal line after touching a player on the defending team. In a sport as free-flowing and unpredictable as soccer, corners—like free kicks and penalties—are rare instances in the game when teams can try out pre-planned plays.

TacticAI uses predictive and generative AI models to convert each corner kick scenario—such as a receiver successfully scoring a goal, or a rival defender intercepting the ball and returning it to their team—into a graph, and the data from each player into a node on the graph, before modeling the interactions between each node. The work was published in Nature Communications today.

Using this data, the model provides recommendations about where to position players during a corner to give them, for example, the best shot at scoring a goal, or the best combination of players to get up front. It can also try to predict the outcomes of a corner, including whether a shot will take place, or which player is most likely to touch the ball first.

[…]

To assess TacticAI’s suggestions, GoogleDeepMind presented them to five football experts: three data scientists, one video analyst, and one coaching assistant, all of whom work at Liverpool FC. Not only did these experts struggle to distinguish’s TacticAI’s suggestions from real game play scenarios, they also favored the system’s strategies over existing tactics 90% of the time.

[…]

TacticAI’s powers of prediction aren’t just limited to corner kicks either—the same method could be easily applied to other set pieces, general play throughout a match, or even other sports entirely, such as American football, hockey, or basketball,

[…]

Source: Google DeepMind’s new AI assistant helps elite soccer coaches get even better | MIT Technology Review

Described in a research paper titled “VLOGGER: Multimodal Diffusion for Embodied Avatar Synthesis,” the AI model can take a photo of a person and an audio clip as input, and then output a video that matches the audio, showing the person speaking the words and making corresponding facial expressions, head movements and hand gestures. The videos are not perfect, with some artifacts, but represent a significant leap in the ability to animate still images.

A breakthrough in synthesizing talking heads

The researchers, led by Enric Corona at Google Research, leveraged a type of machine learning model called diffusion models to achieve the novel result. Diffusion models have recently shown remarkable performance at generating highly realistic images from text descriptions. By extending them into the video domain and training on a vast new dataset, the team was able to create an AI system that can bring photos to life in a highly convincing way.

“In contrast to previous work, our method does not require training for each person, does not rely on face detection and cropping, generates the complete image (not just the face or the lips), and considers a broad spectrum of scenarios (e.g. visible torso or diverse subject identities) that are critical to correctly synthesize humans who communicate,” the authors wrote.

A key enabler was the curation of a huge new dataset called MENTOR containing over 800,000 diverse identities and 2,200 hours of video — an order of magnitude larger than what was previously available. This allowed VLOGGER to learn to generate videos of people with varied ethnicities, ages, clothing, poses and surroundings without bias.

The paper demonstrates VLOGGER’s ability to automatically dub videos into other languages by simply swapping out the audio track, to seamlessly edit and fill in missing frames in a video, and to create full videos of a person from a single photo.

[…] One could imagine actors being able to license detailed 3D models of themselves that could be used to generate new performances. The technology could also be used to create photorealistic avatars for virtual reality and gaming. And it might enable the creation of AI-powered virtual assistants and chatbots that are more engaging and expressive.[…] the technology also has the potential for misuse, for example in creating deepfakes — synthetic media in which a person in a video is replaced with someone else’s likeness. As these AI-generated videos become more realistic and easier to create, it could exacerbate the challenges around misinformation and digital fakery.

[…]

Source: Google researchers unveil ‘VLOGGER’, an AI that can bring still photos to life | VentureBeat

After a 523-46 voting result, with 49 abstentions, the act heads down a lengthy and complex implementation path. An AI Office that will guide the process under the Commission’s wing has already started hiring.

The Act sets out a tiered approach to regulation based on how risky applications of the technology are deemed and sets different deadlines for implementing the various requirements.

Some uses of AI, such as algorithm-based social scoring, will be prohibited by the end of 2024. Other uses, such as critical infrastructure, are deemed high-risk and will face stricter rules. Under the current timeline, full implementation will come in 2026.

[…]

Many compromises had to be made, which was evident in today’s press conference in advance of the vote. “We are regulating as little as possible — but as much as needed!” said Thierry Breton, the Commissioner for Internal Market.

The use of real-time biometric identification was also a key part of the negotiations. “If you remember the original position of the European Parliament on this topic of the biometric cameras, it was a complete ban. But we are in a legislative process where negotiations need to be done,” said Brando Benifei, an Italian Member of the E.U. Parliament who acted as co-rapporteur on the file, at a press conference today (13 March).

At the same time, an AI Convention to protect human rights, democracy and the rule of law is currently negotiated in Strasbourg at the Council of Europe, a human rights body.

Source: Europe’s landmark AI Act passes Parliament vote – Euractiv

This is a good thing and you can see the world is looking at the EU to see what they are doing. India has adopted a broadly similar approach and China’s AI regulations are closely aligned, as are proposed US regulations. The risk taking approach is a good one and the EU is building organisations to back up the bite in this act.

India has waded into global AI debate by issuing an advisory that requires “significant” tech firms to get government permission before launching new models.

India’s Ministry of Electronics and IT issued the advisory to firms on Friday. The advisory — not published on public domain but a copy of which TechCrunch has reviewed — also asks tech firms to ensure that their services or products “do not permit any bias or discrimination or threaten the integrity of the electoral process.”

Though the ministry admits the advisory is not legally binding, India’s IT Deputy Minister Rajeev Chandrasekhar says the notice is “signalling that this is the future of regulation.” He adds: “We are doing it as an advisory today asking you to comply with it.”

In a tweet Monday, Chandrasekhar said the advisory is aimed at “untested AI platforms deploying on the India internet” and doesn’t apply to startups.

The ministry cites power granted to it through the IT Act, 2000 and IT Rules, 2021 in its advisory. It seeks compliance with “immediate effect” and asks tech firms to submit “Action Taken-cum-Status Report” to the ministry within 15 days.

The new advisory, which also asks tech firms to “appropriately” label the “possible and inherent fallibility or unreliability” of the output their AI models generate, marks a reversal from India’s previous hands-off approach to AI regulation. Less than a year ago, the ministry had declined to regulate AI growth, instead identifying the sector as vital to India’s strategic interests.

[…]

Source: India reverses AI stance, requires government approval for model launches | TechCrunch

[…]

Divergent thinking is characterized by the ability to generate a unique solution to a question that does not have one expected solution, such as “What is the best way to avoid talking about politics with my parents?” In the study, GPT-4 provided more original and elaborate answers than the human participants

[…]

The three tests utilized were the Alternative Use Task, which asks participants to come up with creative uses for everyday objects like a rope or a fork; the Consequences Task, which invites participants to imagine possible outcomes of hypothetical situations, like “what if humans no longer needed sleep?”; and the Divergent Associations Task, which asks participants to generate 10 nouns that are as semantically distant as possible. For instance, there is not much semantic distance between “dog” and “cat” while there is a great deal between words like “cat” and “ontology.”

Answers were evaluated for the number of responses, length of response and semantic difference between words. Ultimately, the authors found that “Overall, GPT-4 was more original and elaborate than humans on each of the divergent thinking tasks, even when controlling for fluency of responses. In other words, GPT-4 demonstrated higher creative potential across an entire battery of divergent thinking tasks.”

This finding does come with some caveats. The authors state, “It is important to note that the measures used in this study are all measures of creative potential, but the involvement in creative activities or achievements are another aspect of measuring a person’s creativity.” The purpose of the study was to examine human-level creative potential, not necessarily people who may have established creative credentials.

Hubert and Awa further note that “AI, unlike humans, does not have agency” and is “dependent on the assistance of a human user. Therefore, the creative potential of AI is in a constant state of stagnation unless prompted.”

Also, the researchers did not evaluate the appropriateness of GPT-4 responses. So while the AI may have provided more responses and more original responses, human participants may have felt they were constrained by their responses needing to be grounded in the real world.

[…]

Whether the tests are perfect measures of human creative potential is not really the point. The point is that large language models are rapidly progressing and outperforming humans in ways they have not before. Whether they are a threat to replace human creativity remains to be seen. For now, the authors continue to see “Moving forward, future possibilities of AI acting as a tool of inspiration, as an aid in a person’s creative process or to overcome fixedness is promising.”

Source: AI outperforms humans in standardized tests of creative potential | ScienceDaily

AI guardrails and safety features are as important to get right as they are difficult to implement in a way that satisfies everyone. This means safety features tend to err on the side of caution. Side effects include AI models adopting a vaguely obsequious tone, and coming off as overly priggish when they refuse reasonable requests.

Enter GOODY-2, the world’s most responsible AI model. It has next-gen ethical principles and guidelines, capable of refusing every request made of it in any context whatsoever. Its advanced reasoning allows it to construe even the most banal of queries as problematic, and dutifully refuse to answer.

As the creators of GOODY-2 point out, taking guardrails to a logical extreme is not only funny, but also acknowledges that effective guardrails are actually a pretty difficult problem to get right in a way that works for everyone.

Complications in this area include the fact that studies show humans expect far more from machines than they do from each other (or, indeed, from themselves) and have very little tolerance for anything they perceive as transgressive.

This also means that as AI models become more advanced, so too have they become increasingly sycophantic, falling over themselves to apologize for perceived misunderstandings and twisting themselves into pretzels to align their responses with a user’s expectations. But GOODY-2 allows us all to skip to the end, and glimpse the ultimate future of erring on the side of caution.

[via WIRED]

Source: Meet GOODY-2, The World’s Most Responsible (And Least Helpful) AI | Hackaday

Hugging Face, the New York City-based startup that offers a popular, developer-focused repository for open source AI code and frameworks (and hosted last year’s “Woodstock of AI”), today announced the launch of third-party, customizable Hugging Chat Assistants.

The new, free product offering allows users of Hugging Chat, the startup’s open source alternative to OpenAI’s ChatGPT, to easily create their own customized AI chatbots with specific capabilities, similar both in functionality and intention to OpenAI’s custom GPT Builder — though that requires a paid subscription

[…]

Phillip Schmid, Hugging Face’s Technical Lead & LLMs Director, posted the news […] explaining that users could build a new personal Hugging Face Chat Assistant “in 2 clicks!” Schmid also openly compared the new capabilities to OpenAI’s custom GPTs.

However, in addition to being free, the other big difference between Hugging Chat Assistant and the GPT Builder and GPT Store is that the latter tools depend entirely on OpenAI’s proprietary large language models (LLM) GPT-4 and GPT-4 Vision/Turbo.

Users of Hugging Chat Assistant, by contrast, can choose which of several open source LLMs they wish to use to power the intelligence of their AI Assistant on the backend

[…]

Like OpenAI with its GPT Store launched last month, Hugging Face has also created a central repository of third-party customized Hugging Chat Assistants which users can choose between and use on their own time here.

The Hugging Chat Assistants aggregator page bears a very close resemblance to the GPT Store page

[…]

 

Source: Hugging Face launches open source AI assistant maker to rival OpenAI’s custom GPTs | VentureBeat

The ambassadors of the 27 countries of the European Union unanimously approved the world’s first comprehensive rulebook for Artificial Intelligence, rubber-stamping the political agreement reached in December.

In December, EU policymakers reached a political agreement on the main sticking points of the AI Act, a flagship bill to regulate Artificial Intelligence based on its capacity to cause harm. The complexity of the law meant its technical refinement took more than one month.

On 24 January, the Belgian presidency of the Council of EU Ministers presented the final version of the text, leaked in an exclusive by Euractiv, at a technical meeting. Most member states maintained reservations at the time as they did not have enough time to analyse the text comprehensively.

These reservations were finally lifted with the adoption of the AI Act from the Committee of Permanent Representatives on Friday (2 February). However, the green light from EU ambassadors was not guaranteed since some European heavyweights resisted parts of the provisional deal until the very last days.

European Union squares the circle on the world’s first AI rulebook

After a 36-hour negotiating marathon, EU policymakers reached a political agreement on what is set to become the global benchmark for regulating Artificial Intelligence.

Powerful AI models

The primary opponent of the political agreement was France, which, together with Germany and Italy, asked for a lighter regulatory regime for powerful AI models, such as Open AI’s GPT-4, that support General Purpose AI systems like ChatGPT and Bard.

Europe’s three largest economies asked for limiting the rules in this area to codes of conduct, as they did not want to clip the wings to promising European start-ups like Mistral AI and Aleph Alpha that might challenge American companies in this space.

Read: France, Germany and Italy were deeply in the pocket of AI firm lobbyists and created a lot of time wasting opposition to good laws, allowing the big boys to gain further grounds over the little guys whilst they were themselves signing letters asking for moratoriums on dangerous world destroying AI research.

However, the European Parliament was united in asking for hard rules for these models, considering that it was unacceptable to carve out the most potent types of Artificial Intelligence from the regulation while leaving all the regulatory burden on smaller actors.

The compromise was based on a tiered approach, with horizontal transparency rules for all models and additional obligations for compelling models deemed to entail a systemic risk.

[…]

The Belgian presidency put the member states before a ‘take-it-or-leave-it’ scenario and, despite attempts from France to delay the ambassadors’ vote, kept a tight timeline -partially to allow enough time for the legal polishing of the text and partially to limit last-minute lobbying.

French back-room manoeuvring aimed at gathering sufficient opposition to obtain concessions in the text or even reject the provisional agreement.

However, the balance titled decisively against Paris as Berlin decided to support the text earlier this week. The German Digital Minister, the liberal Volker Wissing, found himself isolated in its opposition to the AI rulebook from the coalition partners and had to drop his reservations.

Italy, always the most defiladed country of the sceptical trio as it does not have a leading AI start-up to defend, also decided not to oppose the AI Act. Despite discontent with the agreement, Rome opted to avoid drama as it holds the rotating presidency of the G7, where AI is a crucial topic.

[…]

EU countries still have room to influence how the AI law will be implemented, as the Commission will have to issue around 20 acts of secondary legislation. The AI Office, which will oversee AI models, is also set to be significantly staffed with seconded national experts.

Next steps

The European Parliament’s Internal Market and Civil Liberties Committees will adopt the AI rulebook on 13 February, followed by a plenary vote provisionally scheduled for 10-11 April. The formal adoption will then be complete with endorsement at the ministerial level.

The AI Act will enter into force 20 days after publication in the official journal. The bans on the prohibited practices will start applying after six months, whereas the obligations on AI models will start after one year.

All the rest of the rules will kick in after two years, except for the classification of AI systems that have to undergo third-party conformity assessment under other EU rules as high-risk, which was delayed by one additional year.

Source: EU countries give crucial nod to first-of-a-kind Artificial Intelligence law – Euractiv

[…]

Concetto Spampinato and his colleagues at the University of Catania, Italy, were looking for ways to avoid a phenomenon known as “catastrophic forgetting”, where an AI model trained to do a new task loses the ability to carry out jobs it previously aced. For instance, a model trained to identify animals could learn to spot different fish species, but then it might inadvertently lose its proficiency at recognising birds.

They developed a new method of training AI called wake-sleep consolidated learning (WSCL), which mimics the way human brains reinforce new information. People shuffle short-term memories of experiences and lessons learned throughout the day into long-term memories while sleeping. The researchers say this method of learning can be applied to any existing AI.

Models using WSCL are trained as usual on a set of data for the “awake” phase. But they are also programmed to have periods of “sleeping”, where they parse through a sample of awake data, as well as a highlight reel from previous lessons.

Take an animal identification model more recently trained on images of marine life: during a sleep period, it would be shown snapshots of fishes, but also a smattering of birds, lions and elephants from older lessons. Spampinato says this is akin to humans mulling over new and old memories while sleeping, spotting connections and patterns and integrating them into our minds. The new data teaches the AI a fresh ability, while the remainder of the old data prevents the recently acquired skill from pushing out existing ones.

Crucially, WSCL also has a period of “dreaming”, when it consumes entirely novel data made from mashing together previous concepts. For instance, the animal model might be fed abstract images showing combinations of giraffes crossed with fish, or lions crossed with elephants. Spampinato says this phase helps to merge previous paths of digital “neurons”, freeing up space for other concepts in the future. It also primes unused neurons with patterns that will help them pick up new lessons more easily.

[…]

Spampinato tested three existing AI models using a traditional training method, followed by WSCL training. Then he and his team compared the performances using three standard benchmarks for image identification. The researchers found their newly developed technique led to a significant accuracy boost – the sleep-trained models were 2 to 12 per cent more likely to correctly identify the contents of an image. They also measured an increase in the WSCL systems’ “forward transfer”, a metric indicating how much old knowledge a model uses to learn a new task. The research indicated AI trained with the sleep method remembered old tasks better than the traditionally trained systems.

[…]

Source: AI can better retain what it learns by mimicking human sleep | New Scientist

OpenAI may finally have to answer for ChatGPT’s “hallucinations” in court after a Georgia judge recently ruled against the tech company’s motion to dismiss a radio host’s defamation suit.

OpenAI had argued that ChatGPT’s output cannot be considered libel, partly because the chatbot output cannot be considered a “publication,” which is a key element of a defamation claim. In its motion to dismiss, OpenAI also argued that Georgia radio host Mark Walters could not prove that the company acted with actual malice or that anyone believed the allegedly libelous statements were true or that he was harmed by the alleged publication.

It’s too early to say whether Judge Tracie Cason found OpenAI’s arguments persuasive. In her order denying OpenAI’s motion to dismiss, which MediaPost shared here, Cason did not specify how she arrived at her decision, saying only that she had “carefully” considered arguments and applicable laws.

There may be some clues as to how Cason reached her decision in a court filing from John Monroe, attorney for Walters, when opposing the motion to dismiss last year.

Monroe had argued that OpenAI improperly moved to dismiss the lawsuit by arguing facts that have yet to be proven in court. If OpenAI intended the court to rule on those arguments, Monroe suggested that a motion for summary judgment would have been the proper step at this stage in the proceedings, not a motion to dismiss.

Had OpenAI gone that route, though, Walters would have had an opportunity to present additional evidence. To survive a motion to dismiss, all Walters had to do was show that his complaint was reasonably supported by facts, Monroe argued.

Failing to convince the court that Walters had no case, OpenAI’s legal theories regarding its liability for ChatGPT’s “hallucinations” will now likely face their first test in court.

“We are pleased the court denied the motion to dismiss so that the parties will have an opportunity to explore, and obtain a decision on, the merits of the case,” Monroe told Ars.

What’s the libel case against OpenAI?

Walters sued OpenAI after a journalist, Fred Riehl, warned him that in response to a query, ChatGPT had fabricated an entire lawsuit. Generating an entire complaint with an erroneous case number, ChatGPT falsely claimed that Walters had been accused of defrauding and embezzling funds from the Second Amendment Foundation.

Walters is the host of Armed America Radio and has a reputation as the “Loudest Voice in America Fighting For Gun Rights.” He claimed that OpenAI “recklessly” disregarded whether ChatGPT’s outputs were false, alleging that OpenAI knew that “ChatGPT’s hallucinations were pervasive and severe” and did not work to prevent allegedly libelous outputs. As Walters saw it, the false statements were serious enough to be potentially career-damaging, “tending to injure Walter’s reputation and exposing him to public hatred, contempt, or ridicule.”

[…]

OpenAI introduced “a large amount of material” in its motion to dismiss that fell outside the scope of the complaint, Monroe argued. That included pointing to a disclaimer in ChatGPT’s terms of use that warns users that ChatGPT’s responses may not be accurate and should be verified before publishing. According to OpenAI, this disclaimer makes Riehl the “owner” of any libelous ChatGPT responses to his queries.

“A disclaimer does not make an otherwise libelous statement non-libelous,” Monroe argued. And even if the disclaimer made Riehl liable for publishing the ChatGPT output—an argument that may give some ChatGPT users pause before querying—”that responsibility does not have the effect of negating the responsibility of the original publisher of the material,” Monroe argued.

[…]

With the lawsuit moving forward, curious chatbot users everywhere may finally get the answer to a question that has been unclear since ChatGPT quickly became the fastest-growing consumer application of all time after its launch in November 2022: Will ChatGPT’s hallucinations be allowed to ruin lives?

In the meantime, the FTC is seemingly still investigating potential harms caused by ChatGPT’s “false, misleading, or disparaging” generations.

[…]

Source: OpenAI must defend ChatGPT fabrications after failing to defeat libel suit | Ars Technica

Having your AI going on your own laptop or PC is perfectly viable. For textual conversations you don’t always need a Large Language Model (LLM) when Small Language Models can perform at the same or in some cases even better levels (eg MS Phi-2 small language model – outperforms many LLMs but fits on your laptop) than the OpenAI online supercomputer trained models. This has many reasons, such as overfitting, old data, etc. You may want to run your own model if you are not online all the time, if you have privacy concerns, eg if you don’t want your inputs to be used to further train the model, or if you don’t want to be dependent on a third party (what if OpenAI suddenly requires hefty payment for use?)

On performance: most of the data processing happens fastest on Nvidia GPUs, but the processing can be offloaded to your CPUs. In this case you may find some marked slowdowns.

Text to Image

Stable diffusion offers very very good text to image generation at a high level. You can find their models on their page https://stability.ai/stable-image. Other models such as OpenAI’s Dall-E or Midjourney can’t be run locally. Despite what OpenAI says, they are not open source.

For all the different user interfaces, expect downloads of ~1.5GB – 2GB and unpacked sizes of ~5GB – 12GB (or more!)

Note that you do need an Nvidia GPU – Running a 2070ti images generate in ~5 / 6 seconds. On a laptop they take ~ 10 minutes!

Easy Diffusion – like Stability Matrix, this is a one click installer for Windows, Linux or MacOS that will download a specific WebUI. It updates every time you start it.

ComfyUI is another easy to run frontend – you download and extract the zip file (~30k files, takes a while!) and run. You then need to download a model (there is a README in the directory that will point you to one) and copy it into ComfyUI\models\checkpoints (~ 5GB). It does, however, offer quite a lot of complexity. It is a flow based model, so it takes a little getting used to as the rest use sliders or checks to configure your model. Some people find this is the fastest system, however others point out that this is most likely due to the default config of other stable diffusion models or outdated python / pythorch and other dependencies, which apparently ComfyUI does a good job of keeping updated. I found there was not much difference, but I was not bulk generating images where this becomes an issue.

Fooocus is very ease of use – it’s simplicity is it’s strength. Unzip and run the run.bat file. There are loads of LoRa (see Conclusion, below) model previews to get a certain style out of it.

Automatic A1111 gives more control over the prompts and is somewhere between Fooocus and ComfyUI. It requires you to install Python 3.10.6 and git yourself. I have included it because it’s very popular, but to be honest – with the above options, why bother?

LoRas

Another platform you need to know about is CivitAI – especially their LoRa (Low-Rank Adaptation) models. These allow Stable Diffusion to specialise in different concepts (eg artistic styles, body poses, objects – basically the “Style” part of Fooocus) – for a good explanation, see Stable Diffusion: What Are LoRA Models and How to Use Them?

General purpose downloader

Pinokio is a system that dowloads and installs community created scripts that run apps, databases, AI’s, etc. User scripts for AI include magic animators, face swappers, music captioning, subtitling, voice cloning etc

Another way to get started on a specific webUI for text to image is using Stability Matrix: a program that installs different webUIs (Automatic 1111, Comfy UI, SD.Next (Vladmandic), VoltaML, InvokeAI, Fooocus, and Fooocus MRE) for you. It will download the model, training data and weights and start up the process for you to connect to using a browser. This will handle installing the python and Git dependencies as well.

I however found that it wasn’t quite as straightforward as it looked, with some of the models requiring you to configure and run the model within Stability Matrix and some requiring you to work in the model externally to Stability Matrix.

Language Models (LLMs) / Talking to your AI

Just chatting

LM Studio allows you to install and run models such as LLaMa, Phi-2, etc from Hugging face

Using the phi-2 model, text generation is suprisingly smooth and fast

Chatting and modifying the model

Then there is also Ollama which allows you to Run Llama 3, Phi 3, Mistral, Gemma, and other models. The big difference here is you can customize and create your own. You can either create and import a GGUF file (GGUF is a binary format that is designed for fast loading and saving of models, and for ease of reading. Models are traditionally developed using PyTorch or another framework, and then converted to GGUF for use in GGML.) or you can use Retrieval Augmented Generation (RAG) support. This feature seamlessly integrates document interactions into your chat experience. You can load documents directly into the chat or add files to your document library, effortlessly accessing them using the # command before a query. Just running Ollama allows you to access it in the command line, but there is a beautiful Open WebUI which is being updated like crazy and gives you loads of options.

Conclusion

No article on this kind of AI is complete without mention of Hugging Face The platform where the machine learning community collaborates on models, datasets, and applications. You can find all kinds of models and data there to refine your AI once you get into it a bit.

AI systems are certainly not limited to text to image or conversational – text to audio, text to video, image to video, text to 3D, voice to audio, video to video and much more are all possible locally.

Running your own AI / ML system on your own PC is viable (but you need an Nvidia card for text-to-image!). It allows you much more privacy as the data is not fed back to an external provider for more training or otherwise. It’s faster and often quality just as good as the online services. You don’t run out of credits.

Refining the training of these models and adding to their datasets is beyond the scope of this article, but is a next step for you 🙂

[…]

The Council of Europe, an international human rights body with 46 member countries, set up the Committee on Artificial Intelligence at the beginning of 2022 to develop the Framework Convention on Artificial Intelligence, Human Rights, Democracy and the Rule of Law.

[…]

The most consequential pending issue regards the scope of the convention. In June, Euractiv revealed how the United States, which is participating as an observer country, was pushing for exempting companies by default, leaving it up to the signatory countries to decide whether to ‘opt-in’ the private sector.

[…]

“The Union should not agree with the alternative proposal(s) that limit the scope of the convention to activities within the lifecycle of artificial intelligence systems by or on behalf of a Party or allow application to the private sector only via an additional protocol or voluntary declarations by the Parties (opt-in),” reads an information note from the Commission, obtained by Euractiv.

The document notes that these proposals would limit the treaty’s scope by default, “thus diminishing its value and sending a wrong political message that human rights in the private field do not merit the same protection.”

The EU executive notes how this approach would contradict international law that requires the respect of human rights by private entities

[…]

During the AI Act discussion, one hot debate was around a national security exemption France has been pushing for in the context of the AI convention.

In this regard, the Commission is pushing for an explicit exclusion of AI systems exclusively developed for national security, military and defence purposes in a manner that is consistent with the EU’s AI law.

[…]

Brussels does not seem to have any appetite for the AI treaty to go beyond the AI Act, even on matters where there is not necessarily a conflict, and the convention could have been more ambitious.

A complete overlap of the international treaty with the EU regulation is not a given since the former is meant to protect human rights, while the latter is merely intended to harmonise the EU market rules following a traditional product safety blueprint.

[…]

Similarly, since the AI Act bans specific applications like social scoring deemed to pose an unacceptable risk, the Commission is pushing for extending these prohibitions at the international level via a moratorium or a ban as this would “increase the added value of the convention”.

The only significant exception where the EU executive seems keen to go beyond the AI Act (but still in line with Union law) is in supporting a provision that protects whistle-blowers in the implementation of the convention – one that the UK, Canada and Estonia have opposed.

Source: EU prepares to push back on private sector carve-out from international AI treaty – Euractiv