Tag: inteligencia artificial

The 5 big differences between ChatGPT and Microsoft Copilot: which one is better?

Do you use artificial intelligence in your daily life? In recent years, AI has experienced a spectacular boom, and its impact is felt in practically every sector. From improving work efficiency to providing new forms of entertainment, AI has transformed the way we interact with technology. In this scenario, two names stand out above others: ChatGPT from OpenAI and Microsoft Copilot, used daily by millions of people due to their impressive capabilities. However, despite ChatGPT and Microsoft Copilot sharing some similarities, such as […]

Do you use artificial intelligence in your daily life? In recent years, AI has experienced a spectacular boom, and its impact is felt in practically every sector. From improving work efficiency to providing new forms of entertainment, AIs have transformed the way we interact with technology. In this scenario, two names stand out above others: ChatGPT by OpenAI and Microsoft Copilot, used every day by millions of people due to their impressive capabilities.

Subscribe to the Softonic newsletter and get the latest in tech, gaming, entertainment and deals right in your inbox.

Subscribe (it's FREE) ►

However, despite the fact that ChatGPT and Microsoft Copilot share some similarities, such as the use of natural language models and the ability to integrate multiple data sources to generate accurate and relevant responses, the differences between these two AIs become quite evident when we delve deeper. Below, we detail the main differences that distinguish ChatGPT and Microsoft Copilot.

Purpose and focus

The first major difference between ChatGPT and Copilot lies in their own purposes. ChatGPT was designed as a generalist conversational AI, so it can answer a wide range of questions and maintain textual and voice conversations that are very varied, from everyday chats to complex technical topics. OpenAI’s AI is designed to be flexible and adapt to almost any context where a fluid dialogue is needed, and its main goal is to generate natural and useful interactions for users in all kinds of situations.

For its part, Microsoft Copilot is more focused on productivity and assisting users with specific tasks within work environments, integrating into programs and tools like Edge, Word, Excel, and PowerPoint to help users with all kinds of tasks more efficiently. Rather than simply maintaining a general conversation, the purpose of Copilot is to automate and simplify processes within a workflow, optimizing tasks that usually consume a lot of time.

Integration with other tools

Another major difference between both AIs is their ability to integrate with other tools. ChatGPT is primarily presented as an autonomous application that users can use directly through a website, an app, or an API. Although it can be integrated with some platforms, its main focus remains conversational interaction without the need to be linked to a specific software suite.

On the contrary, and although it also has its own website, Microsoft Copilot is deeply integrated into the Microsoft ecosystem, allowing it to offer assistance directly in applications like Excel and Outlook. This gives it a huge advantage in work environments where these tools are essential, as Copilot can access the user’s documents and data within these applications to offer personalized suggestions and perform a wide variety of tasks, such as creating charts or automatically drafting emails.

ChatGPT DOWNLOAD

Flexibility versus specialization

The flexibility of ChatGPT contrasts with the specialization of Microsoft Copilot. ChatGPT is capable of adapting to different contexts and is not limited to a specific set of applications or tasks. Users can ask it to explain a mathematical concept, generate a short story, or offer advice on a wide range of topics. This generalist approach allows ChatGPT to be useful for a broad audience, from students to professionals in various fields.

In contrast, Microsoft Copilot focuses on very specific and specialized tasks, particularly within the business sector, and is designed to enhance productivity within platforms like Azure or Microsoft 365, making it a very versatile tool for those who use these in their daily activities. However, although its specialization makes it extremely efficient in these contexts, its usefulness outside of these applications is quite limited.

Interface and user experience

Regarding access to these AIs, ChatGPT is designed to be used through a simple chat interface. Users interact with the AI by typing questions or commands in a chat window, and ChatGPT responds quickly and directly, using the most suitable model for each context. This simplicity makes the user experience very intuitive due to the great similarity to a natural conversation.

For its part, Microsoft Copilot takes a different approach. In addition to using a web chat interface, Copilot integrates directly into the Microsoft ecosystem. This way, by being within programs that the user already knows, such as Word and Edge, the learning curve is reduced and access to its functionalities is facilitated.

Update and customization

Finally, one of the most notable differences is how both systems are updated and customized. ChatGPT receives periodic updates from OpenAI, which expand its knowledge base and improve its ability to generate accurate responses. Additionally, users can customize their interactions through specific prompts or have the AI remember things thanks to its Memory function, allowing it to adapt to different contexts.

Microsoft Copilot, on the other hand, is closely tied to the updates of the programs and systems in which it is integrated, and its ability to be customized largely depends on the user’s data and context within these applications. This makes it a more static tool in terms of its update, as it is limited by the advancements in the Microsoft ecosystem.

Microsoft Copilot DOWNLOAD

OpenAI has just reached a record valuation: No one knows how they have achieved $157 billion so far

Sam Altman’s company has just raised the astonishing amount of $6.6 billion, resulting in a valuation of $157 billion. Even for a huge Silicon Valley company, that’s a lot of money. And amid the company’s downturn, considering the foundational concept of a non-profit organization, OpenAI has not disclosed how it will invest the capital, only stating that it will accelerate AI research and increase its computing capacity. In an unprecedented operation, OpenAI has raised $6.6 billion in new funds, raising its valuation to […]

Sam Altman’s company has just raised the astonishing amount of 6.6 billion dollars, resulting in a value of 157 billion. Even for a huge Silicon Valley company, that’s a lot of money.

ChatGPT DOWNLOAD

In the midst of the company’s downfall, if we consider the foundational concept of a non-profit organization, OpenAI has not disclosed how it will invest the capital, only stating that it will accelerate AI research and increase its computing capacity.

In an unprecedented operation, OpenAI has raised 6.6 billion dollars in new funding, bringing its valuation to the astonishing figure of $157 billion. This investment round is one of the largest in history in private markets and confirms OpenAI’s position as a leader in the AI race.

Billions to Boost the AI of the Future

Thrive Capital, led by Josh Kushner, spearheaded the funding round with an investment of 1.3 billion dollars. Microsoft, the main sponsor of OpenAI, contributed another 750 million dollars, as reported by a source to Bloomberg, which brings Redmond’s total investment to nearly 14 billion dollars.

Other investors included Nvidia, Khosla Ventures, Fidelity Management & Research Co, Tiger Global Management, Altimeter Capital, Coatue Management, and the venture capital firm Quiet Capital.

The funding round also attracted global interest, as the Japanese group SoftBank and MGX, based in Abu Dhabi, joined the list of investors, according to people familiar with the matter who spoke to Bloomberg. SoftBank’s investment amounted to 500 million dollars, according to one of the sources.

Funding has turned OpenAI into one of the three largest venture-backed startups, along with Elon Musk’s SpaceX and ByteDance, the owner of TikTok.

ChatGPT DOWNLOAD

OpenAI plans to use this capital injection to accelerate AI research and expand its computing capacity. “The new funding will allow us to double down on our leadership in cutting-edge AI research, increase computing capacity, and continue creating tools that help people solve difficult problems,” the company stated when announcing the operation.

UK companies are ignoring AI: this latest report bursts the AI bubble

Everyone is obsessed with artificial intelligence. Or at least that’s what it seems if we look at corporate investments over the past two years. Well, according to a new study, British companies continue to prioritize hiring workers with basic technical skills. According to the Indeed study, only 2.6% of job postings in the United Kingdom mention AI skills, while basic skills like Microsoft Office and general IT experience are more common. According to the report, what English companies most frequently request are computer skills […]

Everyone is obsessed with artificial intelligence. Or at least that’s what it seems if we look at company investments over the past two years. Well, according to a new study, British companies continue to prioritize hiring workers with basic technical knowledge.

ChatGPT DOWNLOAD

According to the study by Indeed, only 2.6% of job postings in the United Kingdom mention AI skills, while basic skills like Microsoft Office and general IT experience are more common.

According to the report, what English companies most request are general computer skills (10%), Microsoft Office (6%), and Microsoft Excel (5%). Additionally, the demand for basic computer skills has remained fairly constant over the past five years, both in the United Kingdom and in other markets such as the United States.

More basic knowledge and less AI

In addition to technological skills, Indeed found that UK employers are also prioritizing human skills such as communication (30%), leadership (9%), and organization (7%).

Indeed’s research on the current state of the British labor market addresses the current concern that AI could replace human workers. The analysis of more than 2,800 job skills revealed that it is “very unlikely” or “unlikely” that two-thirds (68.7%) will be replaced by generative AI.

“While AI may eventually require a broad upskilling of the entire workforce to adopt advanced technologies, there remains a more pressing concern around closing basic digital skills gaps and enabling everyone to fully engage with work in the digital age,” the report explains.

ChatGPT DOWNLOAD

The current job market suggests that companies are still not aligned with this vision, as employers continue to seek technological skills and fundamental human competencies.

Not everything has been a success in the world of artificial intelligence: 5 AIs that failed spectacularly

Artificial intelligence has transformed many industries in recent years, such as medicine and entertainment, and revolutionized the way we interact with technology, making our lives much easier. With great examples like OpenAI’s ChatGPT, Google’s Gemini, or Meta’s Llama, AI has achieved impressive advancements, from medical diagnosis to autonomous driving, and has opened up a wide range of possibilities that a few years ago seemed like science fiction. Tech companies have heavily invested in this technology, pouring millions into developing increasingly advanced AI models that learn, interact, and […]

Artificial intelligence has transformed many industries in recent years, such as medicine and entertainment, and revolutionized the way we interact with technology, making our lives much easier. With great examples like ChatGPT from OpenAI, Gemini from Google, or Llama from Meta, AI has achieved impressive advancements, from medical diagnosis to autonomous driving, and has opened up a wide range of possibilities that just a few years ago seemed like science fiction.

Tech companies have heavily invested in this technology, pouring millions into developing increasingly advanced AI models that learn, interact, and solve problems at previously unimaginable levels. However, along this path filled with successes, there have also been stumbles. Quite a few. In fact, some “technological promises” ended up becoming disappointments, whether due to implementation errors, overly high expectations, or simply not being ready for the real world.

Subscribe to the Softonic Newsletter and receive the latest news in tech, games, entertainment in your inbox

Subscribe (it's FREE) ►

Failures show us the limitations of this type of technology and the importance of considering human and social factors when designing AI systems. Below, we review five examples of artificial intelligence that did not achieve the expected success and ended up being forgotten.

Tay, the Microsoft AI that went out of control in just one day

In 2016, Microsoft launched Tay, a chatbot designed to interact with Twitter users, learning and adapting to conversations in real time. The idea was for Tay to mimic the behavior and language of the users it interacted with, demonstrating how AI could learn autonomously. However, what seemed like a promising experiment turned into a disaster in less than 24 hours. Some Twitter users realized that they could influence Tay’s behavior by flooding it with racist and offensive comments, and, in no time, the AI began to replicate those same comments, making extremely controversial statements.

Microsoft immediately withdrew Tay and issued a public apology, admitting that it did not foresee that the AI could be manipulated in that way. The case of Tay highlighted the risks of releasing AI without proper control and how malicious users can exploit its weaknesses. Technology alone is not infallible; the social context in which it is implemented is equally important for its success or failure.

Watson for Oncology: a promise that fell on deaf ears

Watson, created by IBM, generated a lot of buzz in its day by promising a revolution in the health field, especially with its application for oncology. Watson for Oncology was supposed to be a tool capable of helping doctors recommend treatments based on a thorough analysis of medical data, research, and studies. However, the reality was far from that promise. Many hospitals that implemented the system complained that the recommendations were not always accurate and, at times, were based on outdated studies or information not applicable to certain patients.

An article by Stat News revealed that some doctors even distrusted Watson’s suggestions, going so far as to ignore them completely. IBM promised that Watson would transform medicine, but the AI did not live up to expectations, and many of the hospitals that initially adopted it ended up abandoning it. Watson for Oncology is a reminder that, although AI has the potential to change fields as important as medicine, the accuracy and constant updating of data are crucial to its success.

Jibo: the family robot that couldn’t find its place

Launched in 2014, Jibo was an artificial intelligence robot that promised to be a companion for families. This friendly robot could tell jokes, give weather forecasts, and learn from human interactions to offer a more personalized experience. Its design and its expectations captured the attention of many media outlets and technology enthusiasts. However, Jibo never took off as expected.

Although its concept was interesting, Jibo had to deal with several problems. On one hand, its utility was quite limited: beyond basic interaction, it did not offer differentiating features compared to other devices. On the other hand, users criticized its high price and the lack of significant updates. Finally, in 2019, the company closed and Jibo’s servers stopped working, leaving its users with an “inert” robot. Despite everything, Jibo remains an example of how innovation alone does not guarantee success if it is not accompanied by functionality that truly adds value to the user.

ChatGPT DOWNLOAD

Bodega: the AI that tried (unsuccessfully) to replace neighborhood stores

In 2017, two former Google employees launched a project called Bodega, a kind of automated kiosk that aimed to replace neighborhood stores. Equipped with cameras and sensors, these kiosks allowed users to take products and pay automatically without the need for a clerk. However, the idea generated enormous controversy from the beginning. Many people considered the project to be insensitive, as it sought to replace a traditional and social element like neighborhood stores, which are more than just points of sale.

Additionally, many criticized the name “Bodega” for appropriating a culturally significant term for Latino communities in the United States. Following the poor public reception and criticism on social media, the project quickly disappeared. Bodega is a great example of how technological advancements can clash with social and cultural aspects that should not be ignored.

Bixby: the Samsung voice assistant that didn’t convince

Bixby, Samsung’s voice assistant, was introduced to compete with virtual assistants like Siri, Alexa, and Google Assistant. However, Bixby never reached the popularity of its rivals. Although Samsung integrated it into its mobile devices and home appliances, users found that its functionality was rather confusing and, in many cases, inferior to other options available on the market.

Despite Samsung’s efforts to improve Bixby with updates, it never managed to capture the public’s attention in the same way as its rivals. Nowadays, although it is still present on some devices, Bixby has taken a back seat in the company’s strategy, which has started to integrate more popular options.

Bixby Home DOWNLOAD

What is the Goodbye Meta AI movement? We explain the new trend on Instagram and Facebook

The world of AI is in full revolution, and big companies just want to keep gathering content to refine and optimize their artificial intelligences. The issue is that users are already tired of seeing how they train AI with our content. And some have said enough. Or, in this case, Goodbye Meta AI. This message, which has been shared by many users, including celebrities, offers no protection for copyright or privacy, experts explain. It has been shared by actors and sports stars (such as James McAvoy, Ashley Tisdale, and Tom Brady) and […]

The world of AI is in full revolution and the big companies just want to keep gathering content to refine and optimize their artificial intelligences. The issue is that users are already tired of seeing how AI is trained with our content. And some have said enough. Or, in this case, Goodbye Meta AI.

Subscribe to the Softonic newsletter and get the latest in tech, gaming, entertainment and deals right in your inbox.

Subscribe (it's FREE) ►

This message, which has been shared by many users, including celebrities, offers no protection of copyright or privacy, experts explain.

It has been shared by actors and sports stars (like James McAvoy, Ashley Tisdale, and Tom Brady) and hundreds of thousands of other people. But, does it really matter?

Why do people post this on their social media?

Meta, the parent company of the three platforms, has announced that it will proceed with its controversial plans to use millions of Facebook and Instagram posts in the UK to train its artificial intelligence (AI) technology, a practice that violates EU privacy laws.

Meta has said that sharing the story with a “Goodbye Meta AI” does not count as a valid form of objection, just like the previous legal tests periodically shared by users in recent years have not provided them with any of the copyright or privacy protection they expected.

Where does this reactionary movement come from?

According to the fact-checking site Lead Stories, affiliated with Meta, the latest iteration comes from a hoax published on September 1. Its text had some differences but followed a similar pattern:

“Goodbye, Meta AI. I order you to stay away from my personal information and photos, and any private information of mine. A lawyer has advised us to publish this or it could have legal consequences. Since Meta is now a public entity, all members must post a similar statement. If you do not post this at least once, it is assumed that you allow the use of your personal information and photos. I do not give you permission to use any of my data or photos,” it said.

Facebook DOWNLOAD

In more general terms, Snopes reports that this type of post dates back to the early 2000s, when some websites began citing the “code 431.322.12 of the Internet Privacy Act.” Apparently, their owners believed they could immunize themselves against the legal consequences of their online actions by invoking a fictitious law.

Microsoft is going to reopen the most infamous nuclear plant in U.S. history because its AI needs it

One of the biggest trends in the field of computing is the expansion of data centers by large companies, whose Internet services are an indispensable part of our commercial and personal lives. These servers and databases require energy, something that is currently in short supply. And the explosion of AI has exacerbated the problem, with estimates that the global consumption of data centers in the United States will reach 35 GW by 2030: almost double what it was in 2022. It may not seem like much considering that only […]

One of the biggest trends in the field of computing is the expansion of data centers by large companies, whose Internet services are an indispensable part of our commercial and personal lives.

ChatGPT DOWNLOAD

These servers and databases require energy, something that is currently in short supply. And the explosion of AI has worsened the problem, with estimates that the global consumption of data centers in the United States will reach 35 GW by 2030: almost double what it was in 2022.

It may not seem like much considering that the United States alone consumes 4,000 TW of electricity per year, but a large part of energy and environmental planning is based on estimated energy consumption, and the expansion of data centers throws the prediction figures directly into the trash.

Big Tech is turning to nuclear energy instead of using renewables

To maintain a reliable energy base for these centers, the largest tech companies are turning to nuclear energy to provide them with electricity with near-zero carbon emissions.

American companies, such as Amazon, are locating new data centers next to nuclear power plants to take advantage of the uninterrupted and reliable energy that nuclear power offers.

On the other hand, Microsoft has reached a 20-year agreement with Constellation to extract energy from what was the Three Mile Island nuclear power plant, located on the Susquehanna River, in Londonderry Township, near Harrisburg (Pennsylvania), and which is now the Crane Clean Energy Center (CCEC).

On March 28, 1979, the Three Mile Unit 2 reactor experienced a core meltdown due to a strange combination of mechanical failure and human error that caused severe damage to the reactor and the release of radioactive water and iodine due to a failure in the pressurized water cooling system.

Although there were no victims and long-term studies could never confirm a causal increase in cancer cases in the area, Three Mile Island was the worst commercial nuclear accident in the U.S. and arguably the third worst in the world after the subsequent incidents at Chernobyl and Fukushima.

However, the impact of the disaster was swift. It brought the issue of reactor safety to the forefront and became the focus of environmentalists and anti-nuclear lobbying groups. The result was the near-total paralysis of commercial nuclear energy in the United States and many European countries until very recently.

Contrary to popular belief, Three Mile Island was not completely shut down after the accident. The Unit 1 reactor continued to operate until 2019, when it was dismantled for economic reasons. According to the new Microsoft agreement, the restart will be the first of a U.S. reactor after its final shutdown.

The 4 best AIs for generating voice and audio in 2024

Do you have a small project in mind and need the ideal voice? Nowadays, artificial intelligence has advanced so much that anyone can transform simple text instructions into a voice that sounds incredibly natural, with tones and nuances almost indistinguishable from a real human voice. These voice synthesis AIs open up a range of possibilities for businesses, content creators, or simply curious individuals who want to experiment with this technology. From audiobook narration to creating ads or videos on platforms like YouTube, the options are endless. In addition to being […]

Do you have a small project in mind and need the ideal voice? Nowadays, artificial intelligence has advanced so much that anyone can transform simple text instructions into a voice that sounds incredibly natural, with tones and nuances almost indistinguishable from a real human voice. These voice synthesis AIs open up a range of possibilities for businesses, content creators, or simply curious individuals who want to experiment with this technology. From audiobook narration to creating advertisements or videos on platforms like YouTube, the options are endless.

Subscribe to the Softonic newsletter and get the latest in tech, gaming entertainment and deals right in your inbox.

In addition to being a useful resource for content creators, voice generation AIs can enhance accessibility, providing very versatile tools to people with functional diversities that make it difficult for them to communicate traditionally. With different voices and styles to choose from, the future of AI-generated content promises to be more inclusive, dynamic, and customizable than ever. Below, I present to you the four best AIs of 2024 for converting text to audio, both for professionals and amateurs.

ElevenLabs

ElevenLabs has quickly positioned itself as one of the preferred options for those seeking advanced AI in creating realistic voices. This platform is known for its precision and fluidity, allowing the generation of voices with an intonation that seems completely natural. One of its strengths is the ability to customize voices, giving the user the option to adjust the tone, speed, and style of speech, facilitating the creation of a voice that perfectly fits their project. Additionally, it features a highly advanced voice cloning functionality that allows the system to recreate specific voices with an impressive level of realism.

The interface of ElevenLabs is very simple and intuitive, allowing any user to create high-quality audio in a matter of minutes without the need for technical knowledge. Additionally, the developers have maintained a focus on scalability, enabling its integration into enterprise platforms to automate the creation of spoken content. Whether it’s for narrating a corporate video, an audiobook, or simply experimenting with the technology, ElevenLabs is one of your best options.

ElevenLabs ACCESS

Description

Descript is one of the most user-friendly tools for content creators looking for an all-in-one solution for audio and video editing, with the added capability of generating voices from text. Descript has gained popularity for its ‘Overdub’ functionality, which allows users to create custom voices or use one of its pre-trained voices to convert text into audio with a natural sound. This option is ideal for those who need to make quick corrections in recordings or add new segments without having to redo the entire audio.

The approach of Descript is practical and productivity-focused. It not only generates audio, but also allows users to edit the text of a script and automatically adjust the audio to match the changes made, saving a lot of time in the editing process. Additionally, its intuitive and user-friendly interface makes it ideal for both beginners and professionals looking to streamline multimedia content production.

Descript DOWNLOAD

Google Cloud Text-to-Speech

Google is not lagging behind in this race, and its service Google Cloud Text-to-Speech remains a standout option in 2024. This tool is part of the Google Cloud artificial intelligence suite and is one of the most comprehensive on the market. The platform allows converting text into high-quality audio using a wide range of voices, which can be adjusted in terms of speed and tone. Additionally, its neural voice synthesis capability provides very natural and expressive voices, making it an excellent choice for audiobooks, podcasts, and multimedia content in general.

Google has integrated its WaveNet technology, developed by DeepMind, which uses neural networks to generate human voices with incredible realism. This tool is highly flexible and supports multiple languages and dialects, making it an ideal option for global companies that need to adapt to different audiences. Despite its business focus, the interface remains quite accessible, allowing users without technical experience to generate voices with ease.

Google Cloud Search DOWNLOAD

Microsoft Azure Cognitive Services

Microsoft couldn’t be left out of this list, and Azure Cognitive Services presents itself as a very powerful option for those looking for a voice synthesis AI. Azure stands out for the possibility of generating personalized voices through its ‘Custom Neural Voice’ feature. This technology allows users to create a unique and exclusive voice by training the model with specific audio data. This is especially useful for brands that want to create a voice that consistently represents their identity across all their products and services.

Another interesting feature is its integration with other Azure tools, which allows companies to easily incorporate voice generation into their automation processes or virtual assistants. With support for more than 75 languages and variants, Azure Cognitive Services is a versatile option for projects of any scale. Additionally, the service stands out for its security and privacy, offering secure storage and processing functions to protect user data.

Microsoft Azure DOWNLOAD

The United Nations will monitor AI with the highest priority

We all know that AI can be a technology that changes everything for the better, but we also know that it has ample potential to be used in undesirable ways and destroy all the beauty in this world. Its regulation will determine the future of humanity. No half measures. Fortunately for us, a United Nations report published today proposes that an international body oversee the first truly global effort to monitor and govern artificial intelligence. The report, prepared by the Secretary-General’s High-Level Advisory Body on AI, […]

We all know that AI can be a technology that changes everything for the better, but we also know that it has more than enough potential to be used in undesirable ways and destroy all the beauty in this world. Its regulation will determine the future of humanity. No half measures.

ChatGPT DOWNLOAD

Luckily for us, a United Nations report published today proposes that an international body oversee the first truly global effort to monitor and govern artificial intelligence.

The report, prepared by the UN Secretary-General’s High-Level Advisory Body on AI, recommends the creation of a body similar to the Intergovernmental Panel on Climate Change to gather updated information on AI and its risks.

Let all the countries of the world have a say on AI, the challenge of our century

The report calls for a new political dialogue on AI so that the 193 UN members can discuss the risks and agree on measures. It also recommends that the UN take steps to ensure that poorer countries can benefit from AI and contribute to its governance.

These measures include the creation of an AI fund to support projects in these countries, the establishment of AI standards and data exchange systems, and the creation of resources such as training to help countries in AI governance.

Some of the recommendations from the report could be facilitated through the Digital Global Compact, an existing plan to address digital and data gaps between nations.

Finally, it suggests the creation of an AI office within the UN dedicated to coordinating existing efforts within the UN to meet the objectives of the report.

The remarkable capabilities demonstrated by large language models and chatbots in recent years have sparked hopes of a revolution in economic productivity, but they have also led some experts to warn that AI may be developing too quickly and could soon become difficult to control.

Shortly after the appearance of ChatGPT, many scientists and entrepreneurs signed a letter calling for a six-month pause in the development of the technology to assess the risks.

The United Nations wants to address this problem called AI before it’s too late, although many wonder if it isn’t already too late.

The AI paradox: Many young people use it without their parent’s knowledge

It is an open secret that AI is used in almost every area of life. Not just at work. Even the youngest use generative AIs, especially chatbots like ChatGPT, to ask questions, do homework, or seek advice. Interestingly, a survey has found that almost no young person admits this use to their parents. A new study reveals which AI tools teenagers in the United States use and how often, as well as how little their parents know about it. Seven out of ten American teenagers have used […]

It is an open secret that AI is used not just in the workplace, but in almost every aspect of life. Even the youngest ones use generative AIs, especially chatbots like ChatGPT, to ask questions, do homework, or seek advice. The curious thing is that a survey has found that almost no young person admits this use to their parents.

Subscribe to the Softonic Newsletter and receive the latest news in tech, games, entertainment in your inbox

Subscribe (it's FREE) ►

A new study reveals which AI tools teenagers in the United States use and how often, as well as how little their parents know about it.

Seven out of ten American teenagers have used generative AI tools, according to a report published today by Common Sense Media. The nonprofit organization analyzed responses to a survey conducted with parents and high school students between March and May 2024 to assess the scale and contours of AI adoption among teenagers.

How many young people use AI, and why?

More than half of the surveyed students had used text generators and AI chatbots such as ChatGPT and Gemini, as well as search engines with AI-generated results. Around 34% had used image generators like DALL-E, and 22% had used video generators.

A June report on teenagers and AI published by Common Sense Media shows that the pace of adoption is accelerating. In late 2023, less than half of the respondents had used generative AI.

The most common reason for using AI was related to studies; more than half stated that they used it as “homework help,” mainly to “generate ideas.”

The second most common reason was old-fashioned boredom, followed by the translation of content from one language to another. And one in five teenagers had used generative AI tools to joke with their friends.

Almost all AU use is behind the parents’ backs

A notable conclusion from the survey is that many parents have no idea if their children are using generative AI. Only 37% of parents with children who use AI tools knew that they were doing so.

Almost a quarter of parents with children who use AI tools had mistakenly assumed they did not. Most parents had not talked about AI with their children.

Almost half of the parents surveyed were concerned that the arrival of generative AI tools could harm their children’s writing and critical thinking. However, it wasn’t all pessimism; parents were divided on how these tools would affect research skills, and more than a quarter stated that they expected AI tools to help their children generate ideas.

ChatGPT DOWNLOAD

Finally, one of the most concerning points of the study: teachers are more than twice as likely to accuse black students of using generative AI in their assignments compared to their white and Latino peers.

Do you have a small audiovisual project in mind? These are the top 4 AI for creating videos from text

Artificial intelligence is revolutionizing almost every creative field, and the audiovisual world is no exception. Just a few years ago, the process of creating a simple video involved hours, if not days, of work: writing the script, shooting the footage, editing and assembling all the material. Nowadays, with the latest generation of text-to-video AI, all of this can be done with just a few clicks. Is it the future of cinema? Personally, I don’t think so, but it is undeniable that these tools are accelerating digital creativity in amazing ways.

Subscribe to the Softonic newsletter and receive the latest news in technology, games, entertainment in your inbox

Subscribe (it's FREE) ►

Imagine being able to describe a scene with a few lines of text and have an AI turn it into a complete video, with characters, movement, and complex settings. Whether it’s creating visual content for social media, prototypes for a film, or even animations, these AIs greatly simplify the video production process. Although no tool is perfect and there are still limitations such as duration or final quality, the advancements in this field are impressive.

Below, we present the top four AI platforms for generating videos from text in 2024. Each one has its own features and advantages, so you will find options for different needs and creative styles.

Sora

Sora is the latest creation from OpenAI, the company behind ChatGPT. This AI is specifically designed to generate videos from text, a functionality that OpenAI had previously put on the backburner compared to its text and image models. With Sora, you can create complex scenes with realistic characters and movements, from historical settings to futuristic situations, all with a simple prompt.

What sets Sora apart from other AI is its ability to interpret precise details of objects and how they behave in the physical world. The AI can generate characters that express emotions such as sadness or happiness, and even incorporate details such as background props or ambient lighting. In addition, Sora not only creates videos from scratch, but can also fill in missing frames or even extend existing videos.

Despite its achievements, the duration of the generated videos is limited to one minute, and there are still some typical defects of AI generation, such as strange movements or elements that appear out of nowhere. However, OpenAI continues to refine this model, which is already a very versatile tool for any content creator.

Sora (OpenAI) ACCESS

Gen-3 Alpha

Runway is one of the pioneers in AI video generation, and its latest model, Gen-3 Alpha, takes the capabilities of its predecessor, Gen-2, to a new level. This model is able to generate video clips from textual descriptions and images, and it does so at an incredibly fast speed. A five-second video takes only 45 seconds to generate, and a ten-second one takes just 90 seconds.

Gen-3 Alpha stands out for offering full control over the visual aspects of videos, from cinematic style to gestures and emotions of the characters. Moreover, it is great for creating smooth transitions between scenes and creative framing.

Although the duration of the videos is one of its current limitations (maximum 10 seconds), Runway plans to release future versions that will allow the creation of longer and more detailed videos. Its moderation system also ensures that the generated content complies with copyright laws and regulations, and avoids the creation of inappropriate material.

RunwayML ACCESS

Synthesia

Synthesia is a specialized tool for creating educational and commercial videos, using AI to generate presentations with realistic avatars. What sets this AI apart is its ability to generate videos in which non-existent people “speak” any given text, in multiple languages and with almost perfect lip synchronization. This makes it a very popular tool for companies that want to create explanatory videos or tutorials without having to film real actors.

One of the great advantages of Synthesia is its simplicity. You just have to choose an avatar, write the script, select the language, and the AI does the rest. The generated videos can have a duration of up to 30 minutes, making it ideal for corporate or educational presentations. However, although the avatars are quite convincing, they still have a slightly robotic appearance, especially in facial expressions.

Synthesia ACCESS

Pictory

Pictory is another very useful AI, especially for those who work with content for social media or digital marketing. This platform allows you to transform text articles into short videos, with background images and music, ideal for capturing the attention of users on Instagram or YouTube. Pictory’s AI analyzes the content of a text and automatically selects images and clips that fit the theme, in addition to generating subtitles.

The interesting thing about Pictory is that it is designed to be accessible for people without video editing experience. The platform offers a wide range of options to customize the visual style and transition effects, which, combined with its intuitive interface, allows you to create a very attractive video in a matter of minutes. Although it is more focused on short and promotional videos, its automation capability makes it perfect for those who prioritize speed and efficiency.

Pictory ACCESS