The administration of Donald Trump is considering tightening restrictions on the export of advanced artificial intelligence (AI) processors, in a move that could have significant repercussions for the global tech industry and the economies of several countries. According to reports, only entities from the United States and 18 allies, classified as Tier 1 countries, will be able to acquire powerful AI processors without limits. This contrasts with the proposed restrictions for other nations, which could see their access to these critical resources reduced. The new guidelines, framed within the so-called AI diffusion rule, allowed […]
The administration of Donald Trump is considering tightening restrictions on the export of advanced artificial intelligence (AI) processors, in a move that could have significant repercussions for the global tech industry and the economies of several countries.
According to reports, only entities from the United States and 18 allies, classified as Tier 1 countries, will be able to acquire powerful AI processors without limits. This contrasts with the proposed restrictions for other nations, which could see their access to these critical resources reduced.
The new guidelines, framed within the well-known AI dissemination rule, previously allowed Tier 2 countries to obtain up to 1,700 units of Nvidia H100 GPUs without a license.
Nvidia Corp says the DeepSeek AI model is an “excellent AI advancement” that complies with US technology export controls https://t.co/rbC95zC06G
The U.S. does not want anyone else but them to develop AI
However, sources suggest that there is consideration to lower this limit, which could raise concerns among tech companies that fear such restrictions may harm their operations and encourage the development of Chinese alternatives.
The administration is also pressuring Japan and the Netherlands to restrict the maintenance service for semiconductor manufacturing tools in China.
This move could severely impact China’s semiconductor production, while ASML and Tokyo Electron, companies that provide these services, would lose millions in revenue.
In addition, sanctions are being evaluated against Chinese companies such as ChangXin Memory Technologies (CXMT) and Semiconductor Manufacturing International Corp. (SMIC).
However, it is unclear how allies will respond to these proposals or when these regulations will be implemented, as it could take several months before they come into effect, due to the current personnel transition in federal agencies.
Zoom, one of the most used tools in the business world, has introduced a new automation tool powered by artificial intelligence, designed to relieve workers from repetitive and administrative tasks. This innovation will allow employees to focus on activities that generate more significant results as hybrid work becomes established in various industries. The new service, called Workflow Automation, is based on no-code automation templates that users can select and modify according to their needs, thus allowing the integration of third-party applications along with Zoom’s own software. A major advancement based […]
This innovation will allow employees to focus on activities that generate more significant results as hybrid work becomes established in various industries.
The new service, called Workflow Automation, is based on no-code automation templates that users can select and modify according to their needs, thus allowing the integration of third-party applications along with Zoom’s own software.
We just launched a whole bunch of new innovations ? https://t.co/2WvX66yswm Get up to speed ⤵️
1️⃣ Zoom Workflow Automation 2️⃣ Forms for Zoom Docs 3️⃣ Companion audio for Zoom Rooms 4️⃣ Zoom Revenue Accelerator for mobile 5️⃣ Smart note templates for AI Expert Assist ➕ More!… pic.twitter.com/6cBUWYPPyr
Wei Li, head of the company’s Chat Team, emphasized that this tool is designed to be accessible even for those without a technical background, stressing that “empowering people with AI to do more is fundamental to what we do at Zoom.”
Among the features offered by Workflow Automation is the ability to automatically generate summaries after meetings, which can then be added to Zoom, Google, or Word documents, and shared in Team Chat channels, all without the need for additional user intervention after the initial setup.
Zoom has highlighted its no-code approach as an advantage over its competitors, which often require certain technical knowledge to use their tools. The company positions itself as an accessible option for many organizations, which could lead to significant improvements in productivity.
In addition to the new automation tool, Zoom has expanded the functionalities of its AI Companion tool. With the integration of additional microphones through users’ laptops, an improvement in audio quality in Zoom Rooms is promised.
They also announce the launch of a YouTube application within Zoom meetings, ensuring a smoother and lag-free video experience.
Paid subscribers of Zoom will have access to Workflow Automation at no additional cost, as long as they use version 6.3.10 or higher of the desktop application.
Nvidia has revolutionized the world of PC graphics since the introduction of its Tensor Cores in 2017. These hardware units are essential for accelerating artificial intelligence workloads, enabling mixed precision computing and significantly improving gaming performance, especially at high resolutions like 4K and 8K. Tensor Cores are found in all of Nvidia’s advanced graphics and are crucial for technologies like DLSS (Deep Learning Super Sampling). How Tensor Cores work By combining artificial intelligence with native hardware, this innovative processing system provides results that […]
Nvidia has revolutionized the world of PC graphics since the introduction of its Tensor Cores in 2017.
These hardware units are essential for accelerating artificial intelligence workloads, enabling mixed precision computing and significantly improving gaming performance, especially at high resolutions like 4K and 8K.
By combining artificial intelligence with native hardware, this innovative processing system delivers results that surpass standard rasterization, achieving higher frame rates at demanding resolutions.
Since their first generation, they have evolved significantly, with the latest innovations offering up to 30 times more speed compared to their predecessors.
The first generation, launched with the Volta architecture, helped accelerate AI workloads on servers, while the second and third generations, present in the RTX Turing and Ampere series, expanded precision capabilities, introducing new formats that facilitate deep learning and content creation.
The latest evolution, the fifth-generation Tensor Cores in the Blackwell architecture, promises innovations such as multiframe generation, which allows for a qualitative leap in the gaming experience, especially in 8K.
All these improvements not only benefit players; Tensor Cores are also used in workstations for CAD tasks, scientific calculations, and machine learning, demonstrating their versatility beyond the gaming realm.
A professor at the University of Florida, Yuheng Bu, is developing an innovative invisible watermark technology designed to detect content generated by artificial intelligence (AI). This technology aims to address the growing challenge faced by academics and employers: discerning whether a text was written by a human or by AI. A study conducted in the UK revealed that 94% of AI-generated texts went undetected, highlighting the urgency of finding effective solutions. The need to identify AI-generated text Bu’s proposal involves that […]
A professor at the University of Florida, Yuheng Bu, is developing an innovative invisible watermark technology designed to detect content generated by artificial intelligence (AI).
This technology aims to address the growing challenge faced by academics and employers: discerning whether a text was written by a human or by AI.
In a study conducted in the United Kingdom, it was revealed that 94% of AI-generated texts went undetected, highlighting the urgency of finding effective solutions.
Think you can cheat with AI? A UF professor creates watermarks to detect AI-generated writing – A University of Florida engineering professor is developing this technology right now. “If I'm a student and I'm writing my homework with ChatGPT, I … – https://t.co/L66K9NhQAR
Bu’s proposal implies that the entity applying the watermark also possesses the key necessary for its detection, raising questions about intellectual property and access to such keys. According to Bu, the quality of the text generated by AI will be maintained after applying the watermark, ensuring that it is robust against modifications such as paraphrasing and synonym replacement.
Working with watermarks could become a crucial tool to ensure trust and authenticity, facilitating verification in academic and digital environments, and helping to differentiate genuine content from misinformation.
This claims that its method offers an improvement over other developments, such as those from Google DeepMind, by applying watermarks only to a subset of text during generation.
However, a significant challenge lies in how end users obtain that watermark key. Currently, a teacher must contact the entity that has embedded the watermark to access the key, raising questions about the control and ethical implications that this approach generates.
In this regard, Bu emphasizes the need to establish an ecosystem that ensures the effective use of watermarks and the distribution of keys in the future.
A new industry report has revealed that both audiences and journalists are increasingly concerned about the impact of generative artificial intelligence (AI) on journalism. Led by Dr. T.J. Thomson from RMIT University in Melbourne, Australia, the report “Generative AI & Journalism” highlights that the potential for AI-generated or edited content to misinform and deceive is the primary concern. According to the study, many journalists are not equipped to identify AI-generated or edited content, leaving them vulnerable to unknowingly spreading this information. This is due to […]
Directed by Dr. T.J. Thomson from RMIT University in Melbourne, Australia, the report “Generative AI & Journalism” highlights that the potential of AI-generated or edited content to misinform and deceive is the main concern.
According to the study, many journalists are not equipped to identify AI-generated or edited content, leaving them vulnerable to spreading this information unknowingly. This is partly because few newsrooms have systematic processes in place to verify community-generated visual material.
“Generative AI can assist our journalists in uncovering the truth and helping more people understand the world. Machine learning already helps us report stories we couldn’t otherwise, and generative AI has the potential to bolster our journalistic capabilities even more.” pic.twitter.com/v2cOuwJzrh
Most of the interviewed journalists were not aware of the growing and often invisible integration of AI in cameras and image and video editing software.
While only a quarter of news audiences reported having encountered generative AI in journalism, a significant percentage expressed uncertainty about the use of this technology.
This lack of clarity reflects concerns about the transparency of news organizations, which could affect public trust in the content they consume.
The report also notes that there are known and unknown biases in AI algorithms that harm different demographic groups. These biases are related to the incorporation of human prejudices in the training data of AI.
Among the most relevant fears is the possibility that AI will replace human work in newsrooms, which could reduce opportunities and skills in the journalism sector.
Finally, the study suggests that AI tools that offer greater transparency and explain their decisions are less risky for journalists compared to those that lack these features.
A recent report from Fortanix reveals that 97% of companies block or restrict the use of Generative AI (GenAI), while 89% of surveyed executives believe these control measures are effective. Despite these restrictions, 95% of professionals are already using some form of artificial intelligence, and a notable 66% employ GenAI in their work, with 64% accessing these tools through personal email accounts. The report, which is based on the opinions of 1,000 executives, also notes that the pressure to adopt GenAI is palpable. 74% of […]
A recent report from Fortanix reveals that 97% of companies block or restrict the use of Generative AI (GenAI), while 89% of surveyed executives believe that these control measures are effective.
Despite these restrictions, 95% of professionals are already using some form of artificial intelligence, and a notable 66% are employing GenAI in their work, with 64% accessing these tools through personal email accounts.
The report, which is based on the opinions of 1,000 executives, also notes that the pressure to adopt GenAI is palpable. 74% of executives feel they must implement these solutions imminently, and 88% have budget allocated for it. This pressure mainly comes from senior management, boards of directors, and market competition.
✅ 97% of companies plan to adopt GenAI ❌ 87% of security execs reported a breach last year ? Most companies restrict GenAI, yet employees still use it—on personal accounts Outdated security won’t cut it. Get the Fortanix 2025 GenAI Security Report https://t.co/QxqEksqUd6pic.twitter.com/wOEp1KU8gr
However, this interest in the implementation of GenAI occurs in a context of significant concerns about data security. According to the report, 87% of security executives reported having faced a data breach in the past year.
Despite this, both business and IT executives (82% and 81%, respectively) believe that their company must advance in the use of GenAI, surpassing security executives who see this need in 56% of cases.
Anuj Jaiswal, product director at Fortanix, points out that while many organizations do not fully understand the complex data security issues related to this technology, it is essential to move forward with confidence.
Fortanix proposes that, through a unified and scalable platform, organizations can manage encryption and password management from a single interface.
Regarding the main concerns about the implementation of GenAI, both business executives and those from IT and security agree that the accuracy of the models is one of the main concerns, with IT executives also prioritizing data security and privacy as a key issue.
La inteligencia artificial (IA) está revolucionando el campo creativo, generando un vórtice de controversias en Hollywood. Los actores y guionistas han expresado su rechazo a la posible explotación de su imagen y talento sin su consentimiento o compensación. Robert Downey Jr., conocido por sus icónicas interpretaciones como Iron Man, ha surgido como uno de los … Continúa leyendo “”
Robert Downey Jr., conocido por sus icónicas interpretaciones como Iron Man, ha surgido como una de las voces críticas más fuertes respecto al uso de la IA en la industria del cine.
En un reciente podcast, Downey Jr. expresó su preocupación por la posibilidad de que su imagen fuera recreada utilizando tecnologías de IA y deepfake.
Robert Downey Jr. says "I intend to sue all future executives" who create a digital replica of him through AI or deepfake technology. And when he's dead? "My law firm will still be very active."
“The idea that somehow [AI] belongs to them because they have these super huge… pic.twitter.com/RaBnLtHbPM
The actor warned about the threat that this practice poses to the identity and work of performers, insisting on the need to establish a contractual language that safeguards actors from being impersonated without authorization.
“Artificial intelligence poses a threat to creative professions,” stated the actors’ union, highlighting the dangers associated with this technology.
Mientras que algunos actores están abiertos a la idea, la mayoría es escéptica y protectora de su legado profesional.
Scarlett Johansson has also been a notable advocate for the protection of artistic image, having faced issues related to the unauthorized use of her voice and image.
La creciente preocupación entre las partes interesadas ha llevado a un llamado a la acción y a la búsqueda de regulaciones legales que rijan el uso de la IA en la industria del entretenimiento, un tema que, aunque aún no parece urgente, podría transformar radicalmente la forma en que se desarrolla la producción cinematográfica en el futuro.
Since their launch in October 2023, the Ray-Ban Meta smart glasses have experienced remarkable success in the market, reaching over 2 million pairs sold. This progress is reinforced by the recent announcement from Meta’s CEO, Mark Zuckerberg, who stated that sales have reached one million pairs just in the first quarter of 2024. With this growing sales trend, EssilorLuxottica, the main partner behind the product, is planning to increase its production to 10 million units annually by the end of 2026. New glasses and a target of 10 million to boost […]
This progress is reinforced by the recent announcement from Meta’s CEO, Mark Zuckerberg, who stated that sales have reached one million pairs just in the first quarter of 2024.
With this growing sales trend, EssilorLuxottica, the main partner behind the product, is planning to increase its production to 10 million units annually by the end of 2026.
Ray-Ban Meta glasses have hit 2 million sales, with production set to ramp up big timehttps://t.co/sowQsSWG96
Francesco Milleri, CEO and President of EssilorLuxottica, has stated that these glasses are intended to become the primary digital platform to meet everyday needs, viewing it as a “shared platform” that invites collaboration with other brands.
Rumors suggest that Meta may be developing a new version of its smart glasses that would include a small display. This innovation could open up new possibilities, such as the inclusion of additional features and visual signals, thus expanding the functionality of the glasses.
Currently, this device is controlled through touch and voice functions, so incorporating a screen could significantly enhance the user experience.
Additionally, Milleri hinted that all future smart glasses from Meta will be powered by artificial intelligence, and the implementation of subscription services is being considered. This aligns with EssilorLuxottica’s vision of bringing Meta’s multimodal technology to a global audience.
The Chinese appliance manufacturer Skyworth has announced the launch of a new smart TV that will incorporate DeepSeek technology, a low-cost language model developed in China. This advancement in artificial intelligence has had a significant impact on the industry, evidenced by the loss of more than $500 billion in market value for Nvidia, the well-known chip manufacturer, which has seen an increase in interest for more affordable chips among companies. Skyworth will integrate DeepSeek into its Kukai AI OS operating system, allowing the TV to recognize local dialects and understand […]
The Chinese appliance manufacturer Skyworth has announced the launch of a new smart TV that will incorporate DeepSeek technology, a low-cost language model developed in China.
This advance in artificial intelligence has had a significant impact on the industry, evidenced by the loss of more than 500 billion dollars in Nvidia’s market value, the well-known chip manufacturer, which has seen an increase in interest for more affordable chips among companies.
Skyworth will integrate DeepSeek into its Kukai AI OS operating system, allowing the television to recognize local dialects and understand vague and ambiguous contexts.
The new "#AI Search" feature now includes DeepSeek-R1's deep thinking service. #Tencent announced on Feb 16 that currently some test users can access it for free, with a full rollout coming soon. pic.twitter.com/bxDBNCJTMv
This feature is crucial in a country like China, where several languages and dialects coexist. Thanks to the semiotic understanding of the model, users will be able to interact with their television in their local dialect, thus improving accessibility and user experience.
The new model, known as Skyworth G7F Pro, promises to be versatile and adapt to various needs. In addition to verbal interaction, the television is expected to generate ambient music, support interactive painting, act as an AI oral companion, create illustrated books, and even plan family trips, including itinerary bookings and reminders.
This wide range of features aims to position the G7F Pro not just as a television, but as a multifunctional tool for the modern home.
However, it is still unknown whether the television will require an internet connection to access all these features. It is possible that DeepSeek will run locally if the device is equipped with a suitable NPU chip.
This innovation demonstrates how the technology sector is evolving towards more accessible and effective solutions, providing consumers with increasingly personalized options in their daily lives.
One of the heaviest arguments from die-hard fans of Artificial Intelligence is that in a few years it will be able to make your greatest dreams come true: you will become Luke Skywalker, it will create endless sequels of Avengers, you will have your personalized series finale, exclusive, just for you, tailored to the parameters you enjoy the most. Do you want Phineas and Ferb to appear in Game of Thrones and for you to help them get crowned? Of course you do. You will be able to see all that and more by entering a prompt. And I won’t even deny that it’s possible, because after all the […]
One of the heaviest arguments from die-hard fans of Artificial Intelligence is that in a few years it will be able to make your greatest dreams come true: you will become Luke Skywalker, it will create endless sequels of Avengers, you will have your personalized, exclusive series finale, just for you, tailored to the parameters you enjoy the most. Do you want Phineas and Ferb to appear in Game of Thrones and help them get crowned? Of course you do. You will be able to see all that and more by just entering a prompt. And I won’t even deny that it’s possible, because after all, the world has embraced generative AI, with all its mistakes, its junk, and its nonsense, with a strong hug that says “Thank you, we can finally stop paying those annoying creatives.” But is that really what we want? Really?
Audiovisual content, especially since the advent of streaming, is an individual activity that becomes collective when we talk about it. We don’t watch series just for our own enjoyment, but to discuss them over coffee and recommend them to our friends. The same goes for movies: enjoying them (or not) is just a small part of the process, but it usually doesn’t stop there. Then comes the moment to share your opinion on platforms like Letterboxd and social media, see what others have thought, modify your point of view, and evolve as a viewer.
Have you ever talked about a movie that you didn’t like and ended up appreciating why other people value it positively after a conversation or reading reviews online? It is part of our audiovisual training that helps us build our own opinion, which can vary based on your experience and tastes. Our criteria grow and evolve because it is part of a shared experience, because you can read what others think about what you just watched, because you can modify your opinion based on others.
That’s why I don’t fully understand -nor will I ever understand- the excitement of some for having an absolutely personalized experience, beyond the gimmick, the carnival attraction, the mere curiosity. Would the ending of Lost be better if you were the screenwriter? Would you really enjoy it if there were no room for surprise because everything is tailored to your interests? And above all, do you really want to see something that you can’t discuss with anyone because it belongs only to you? I don’t think so.
The churro factory
The main function of AI is to please you always. Literally, you are right constantly, even if that means lying to you or giving you incorrect data. Of course, it will not dare to say that the fantastic idea you had makes no sense or that the dramatic progression would be completely broken: in a world where we have been increasingly accustomed to series and movies that repeat the plot over and over so that people looking at their phones do not feel excluded, it is likely that a script written in the moment by AI could satisfy a percentage of the population.
But in the end, these kinds of ideas are just part of the circus, the fair attraction, of sharing on social media “what you got.” If we are already overwhelmed with the thousands of series and movies that come out each year that interest us, are we really going to spend time on a personalized experience that we can’t talk about with anyone afterwards because they won’t understand us, nor will they be interested? Imagine for a moment that the ending of Lost was different and modified for each person: there would have been no controversy, and it wouldn’t still be talked about 15 years later.
Damn, we wouldn’t have even been lost, focused on our own individual navel-gazing, where no one challenges us, no one looks us in the eye and holds a mirror in front of us, they only live to please us. Art must confront you with yourself, with your own fears, with gray characters, with things you would never believe you would enjoy. And for that, it needs the human hand, not that of a machine which, at best, can create a basic coherent story. If your dream is for Spider-man to take off his mask and for you to appear in the movie, what you want is not to see a movie, but the typical Christmas video you used to send to your family in proto-Internet where some elves danced and had your face.
If we normalize that AI takes care of our leisure, they have won. If we want that to be the case, we have lost. Deep down, it is that basic. It is that sad.
.png){kind=icon}
