Technology - Waltes

Yann LeCun’un Vizyonu: Gelecekteki Yapay Zeka Nasıl Olacak?

Yann LeCun, yapay zeka dünyasının önde gelen isimlerinden biri olarak, yeni bir AI sistemi üzerinde çalışıyor. Meta’nın eski baş AI bilim

Yann LeCun, yapay zeka dünyasının önde gelen isimlerinden biri olarak, yeni bir AI sistemi üzerinde çalışıyor. Meta’nın eski baş AI bilimcisi olan LeCun, 2025 yılında bu şirketten ayrılarak Advanced Machine Intelligence Labs (AMI Labs) adlı bir firma kurdu. Hedefi, ChatGPT, Claude ve Gemini gibi mevcut sistemlerin ötesine geçmek. LeCun, bu mevcut sistemlerin belirli işlevleri yerine getirebildiğini kabul ediyor ama gerçek dünya gibi karmaşık durumlarla başa çıkacak yeteneklere sahip olmadığını vurguluyor. “Gerçekten zeki değiller,” diyor LeCun, VivaTech etkinliğinde bana söylediklerinde. Gerçekten de, büyük dil modelleri (LLM’ler) sadece bilgiyi biriktiriyor ve onu geri veriyor. “Onları eğitiyorsun, ama aslında temel bir anlayışları yok,” diye ekliyor.

Gelelim robotlara… LeCun, insansı robotların gerçek dünyayı anlaması için daha esnek bir yapay zekaya ihtiyaç duyulduğunu belirtiyor. Robotik endüstrisine milyarlarca dolarlık yatırım yapılmasına rağmen, bu robotları güvenli bir şekilde ev işlerini yapabilmeleri için eğitmek oldukça zor ve maliyetli. “LLM’ler, robotlar için büyük ölçüde işe yaramaz,” diyor LeCun. Yani, süper zeka seviyesine ulaşmamız pek mümkün değil. Gelecek on yılın, açıklama yapabilen sistemlerin geliştirilmesiyle geçeceğini düşünüyor. “Ne önemli? Ne neden olur? Farklı bir eylem yaparsam ne olur?” gibi sorulara yanıt verebilen modellerin gerekliliğine işaret ediyor.

LeCun, bilgiye organize bir şekilde ulaşabilen sistemlerin önemini vurguluyor. “Bilinçli bir mekanik dünya modeline ihtiyacınız var,” diyor. Bu tür sistemler, bilgiyi hatırlamak, birleştirmek ve gerektiğinde değiştirmek için yapılandırılmış olmalı. Geçmişte, 2017 veya 2018 yıllarında, ChatGPT benzeri bir şeyin ne zaman olacağını tahmin edenler, bunun on yıllar alacağını söylerdi. Ama şimdi, eğer bu sistemler başarılı olursa, büyük düşünme zamanı gelecek. Peki, robotlar bağımsız bir şekilde çalışmaya başladığında insanlara ne olacak? “Yapay zeka bizim için çalışacak,” diyor. Bu, düşündürücü bir soru. Bakalım bu sürecin nereye gideceği…

Kaynak: Orijinal Haber

EU Data Protection Rules Delay AI Tools Rollout: A Growing Concern

Global tech companies are increasingly pointing fingers at the European Union’s regulatory framework as the main culprit behind the delays and non-

Global tech companies are increasingly pointing fingers at the European Union’s regulatory framework as the main culprit behind the delays and non-rollouts of their latest services in Europe. It’s a familiar tune for many European consumers who find themselves waiting longer than their American counterparts for the latest tools and technologies. A recent report by Goverce.AI, analyzing 375 large language models (LLMs) released between June 2018 and May 2026, paints a stark picture of the situation. The findings reveal that at least 11% of model releases from big players like Meta, Google, OpenAI, and Anthropic faced delays or were completely withheld in the EU, in comparison to just 7% in the UK.

Now, hold on a second. Out of the 68 documented cases of delays and non-releases, regulatory factors were the primary reason in a staggering 56 cases. One striking example is the Claude 3 Opus web app, which suffered a 71-day delay in its release in the EU. And guess what? Meta took the cake with the highest overall rate of delays, as more than a quarter—26%—of its releases were either delayed or not launched at all in the EU. In the UK, this number was a bit more forgiving at 15%.

The report highlights that data protection regulations are the main regulatory barrier for these companies. Non-text modalities like images, audio, and real-time video face even tougher challenges than plain old text. With the Digital Markets Act (DMA) only coming into force in 2023 and the AI Act set to be adopted in 2024, the full impact of these regulations is still on the horizon.

Now, while the UK and EU might share a similar regulatory landscape with laws like GDPR, which was in place before Brexit, it seems the EU’s approach is a bit more stringent. The authors of the report argue that this is largely due to the EU’s “aggressive enforcement” and the sluggish pace at which data protection rules are clarified when it comes to training and deploying LLMs. It’s like the EU knows its own rules could be stifling AI development, as they’re currently mulling over the Digital Omnibus, aimed at making data regulations more adaptable for AI growth.

But, wait! There’s another layer to this. The EU is also in the midst of reviewing the EU Copyright Directive alongside the AI Act’s copyright provisions to ensure authors’ rights are protected. This could complicate matters further, making the availability of state-of-the-art AI models even trickier down the line if these regulations are enforced too stringently.

John Lidiard, an AI policy researcher at GovAI, sums it up: the EU’s tight grip on data protection could turn into a chokehold for innovation. This story first made waves on EU Tech Loop and has since been shared on Euronews as part of an agreement.

So, what’s next for AI in Europe? Will these regulatory hurdles continue to slow down innovation, or will there be a breakthrough? Only time will tell…

Kaynak: Orijinal Haber

Meta Glasses Users Face Paywall for Key Features: A Controversy Unfolds

Meta has stirred up a storm among its smart glasses users by putting a crucial feature behind a paywall. Owners of MetaConversation Focus, the featur

Meta has stirred up a storm among its smart glasses users by putting a crucial feature behind a paywall. Owners of MetaConversation Focus, the feature in question, will now face limitations if they use it for more than three hours a month. Meta claims that those who exceed this free monthly usage limit will need to subscribe to a premium plan to continue using the service. Many users have expressed their discontent, saying that putting such a popular feature behind a paywall feels completely unjust. One frustrated user said, “I would gladly subscribe to Meta One, but only if it genuinely offers unlimited access.”

The subscription model is designed for power users who are looking for expanded access and additional perks, such as premium device support. However, this move has left many scratching their heads, especially since Meta has been pushing its Ray-Ban smart glasses — among the most popular devices of their kind — which already face stiff competition from brands like Snapchat and Apple. Recently, Meta expanded its partnership with the Italian eyewear brand to launch its own line of Meta Glasses, with prices starting at £269 in the UK and $299 in the US.

But it’s not just the pricing structure that’s causing waves. Critics have raised eyebrows about privacy concerns related to smart glasses. Some users are claiming that these devices represent an invasion of privacy, and regulators have even contacted Meta regarding reports of workers viewing sensitive videos recorded through the AI glasses. The backlash is palpable, especially since Meta’s decision to invest in smart glasses comes at a time when their metaverse ventures seem to be in question.

As the tech landscape shifts, the question remains: will users be willing to pay for features that were once free? The journey of Meta Glasses continues to unfold, and it seems like this is just the beginning of the controversy.

Kaynak: Orijinal Haber

Google Faces €4.1bn Fine for Anti-Competitive Practices in Android

Google is in hot water again, folks! The European Union has slapped the tech giant with a whopping €4.1 billion fine, claiming that its practices c

Google is in hot water again, folks! The European Union has slapped the tech giant with a whopping €4.1 billion fine, claiming that its practices concerning Android systems have been blocking competition and stifling innovation. This isn’t the first time Google has faced such penalties; back in 2018, the company was accused of three major illegal activities related to its Android operating system.

First off, they allegedly required manufacturers of Android devices, like handsets and tablets, to pre-install Google Search and Chrome as a condition for accessing the Google Play Store. Can you believe that? It’s like saying, “If you want to sell my products, you have to use my services.” Secondly, Google was reportedly making payments to major manufacturers and mobile operators who agreed to only pre-install the Google Search app. Talk about a monopoly move!

And it doesn’t stop there. The allegations also include that Google prevented these manufacturers from selling any smart devices that were running on alternative versions of Android—what they call “forked” versions. But here’s the kicker: it’s important to note that Google’s policies don’t stop device owners from downloading other browsers or using different search engines. They’ve got a point there, but the EU seems to think otherwise.

Google has staunchly rejected the accusations, arguing that their business model for Android has actually created more options for consumers—not less. So, who’s right? The matter is complex, and it’s clear that the tech landscape is constantly evolving. The ramifications of this fine could ripple through the industry, affecting not just Google but also the manufacturers and developers who rely on Android.

In the big picture, this fine is a significant amount, roughly equivalent to a small nation’s GDP! It’s a stark reminder of the ongoing battle between tech giants and regulatory bodies trying to ensure fair competition.

So, what’s next for Google? Will they appeal the fine or take steps to change their practices? Only time will tell, but one thing’s for sure—the eyes of the tech world will be on them.

Kaynak: Orijinal Haber

How AI is Transforming Post-Earthquake Rescue Efforts in Venezuela

In a significant move to enhance rescue operations following devastating earthquakes in Venezuela, organizations like NASA, Microsoft, and the Europe

In a significant move to enhance rescue operations following devastating earthquakes in Venezuela, organizations like NASA, Microsoft, and the European Union are leveraging artificial intelligence (AI) in groundbreaking ways. These entities are employing AI technologies to streamline and expedite rescue efforts, ensuring that aid reaches those in dire need as swiftly as possible.

After a major quake, the chaos and destruction left in its wake often complicate rescue missions. However, with the integration of AI, these organizations are not only improving response times but also increasing the efficiency of their operations. For instance, NASA has been utilizing AI algorithms to analyze satellite imagery, allowing them to identify the most affected areas and prioritize where resources are needed the most. Imagine this: a satellite in orbit captures real-time images of the destruction, and within moments, AI processes this data to highlight critical zones…

Meanwhile, Microsoft is contributing by enhancing communication systems through AI-driven tools. These tools help coordinate between various rescue teams, ensuring that information is shared seamlessly. This is vital, especially when traditional communication methods fail due to infrastructure damage. The ability to communicate effectively can mean the difference between life and death in these situations. The tech giant has also developed AI applications that can assist in locating survivors using data from social media and other digital footprints. You wouldn’t believe how advanced this tech is!

On the other side of the Atlantic, the EU is providing funding and support for AI initiatives that focus on disaster response. Their goal is clear: to foster innovations that can be rapidly deployed during emergencies. This includes AI-powered drones that can survey disaster sites, delivering critical data back to rescue teams on the ground. These drones are not just flying around aimlessly; they are equipped with sensors and cameras that can detect heat signatures, potentially locating trapped individuals beneath rubble. Can you imagine the hope that brings?

But let’s not forget the human element here. As these organizations implement these advanced technologies, they are also training local teams in Venezuela to effectively use AI tools. The idea is to empower communities, giving them the skills needed to respond to emergencies more efficiently. This is a game-changer. Local volunteers and rescue workers are becoming the first line of defense, armed with knowledge and technology that can save lives.

In conclusion, the combination of AI technology and human effort is creating a powerful alliance in the fight against the aftermath of earthquakes. The focus on quick, informed responses can dramatically alter the landscape of disaster management in Venezuela. As these efforts continue to evolve, one has to wonder: what other innovations lie ahead that could further revolutionize disaster response?

Kaynak: Orijinal Haber

Experts Warn: Vehicle Trackers May Not Retrieve Stolen Cars!

Ian Fogg’s experience with his stolen car is a stark reminder that vehicle trackers might not be the safety net many believe them to be. Despite hav

Ian Fogg’s experience with his stolen car is a stark reminder that vehicle trackers might not be the safety net many believe them to be. Despite having real-time location data and video footage of the theft, Fogg has been unable to recover his vehicle. This incident sheds light on a growing concern voiced by experts: car safety may be more about convenience than actual security.

Fogg shared his ordeal with BBC News, highlighting how easy it seems for thieves to steal cars that cost significantly more than a smartphone but utilize similar radio technology. He witnessed his car being driven away through his video doorbell – a moment that must have been terrifying. Initially, he tracked his car using an Apple Airtag hidden inside it. However, the thieves quickly discovered it, discarding the device because it was making noise – a feature Apple added to deter stalking.

When Fogg reached out to Kia Connect for assistance, he found the process frustrating. Each time he requested the location of his car, he was required to fill out a form. After submitting the form eight times, he still didn’t receive any location information until 24 to 48 hours later. The last known location of his vehicle was in Lithuania, but by then, it was already too late.

This situation raises serious questions about the effectiveness of vehicle tracking systems. According to Thatcham Research, a car safety firm, there’s a “genuine and growing gap” between what consumers expect and what these devices can actually deliver. While both cars and smartphones are equipped with mobile phone radios and satellite location chips, the process of tracking stolen vehicles is often bogged down by regulatory requirements, including GDPR compliance, which can delay action and add complexity to the recovery process.

Thatcham Research advises vehicle owners to invest in tracking devices that are independently certified as dedicated stolen vehicle tracking products. These devices should have their own power source and provide genuine real-time location monitoring via a professional monitoring center. The reality is sobering: in the UK alone, nearly 55,000 cars were stolen in 2025, and only about 13% of those vehicles are recovered. This situation not only highlights the vulnerabilities of current tracking technology but also the need for better solutions in protecting vehicles.

Fogg’s story serves as a cautionary tale for car owners everywhere. It’s a stark reminder that relying solely on technology may not be enough to safeguard your vehicle. With theft rates still alarmingly high, what’s next for those who invest in security features? Will technology ever catch up with thieves?

Kaynak: Orijinal Haber

UNICEF’in Raporu: 20 Milyon Çocuk Yapay Zekayı Benimsedi!

At least 20 million children across 10 countries have dived into the world of artificial intelligence, and guess what? They’re adopting it way fast

At least 20 million children across 10 countries have dived into the world of artificial intelligence, and guess what? They’re adopting it way faster than adults! The new analysis from UNICEF reveals that children are outpacing their older counterparts when it comes to using this technology. It’s a real eye-opener. More than 2 million kids, that’s one in every ten, said they turn to AI for advice on things that keep them up at night. Can you believe it?

UNICEF pointed out that kids are more exposed to AI systems, from how these systems are designed to the sneaky ways their personal data is used. Yet here’s the kicker: they have way less power to avoid or challenge these systems. Most AI governance just doesn’t prioritize children at all. It’s pretty alarming, right? The agency emphasized that we’re just starting to understand how AI affects cognitive development, emotional dependency, and potential harm to young minds.

And it gets even more concerning. Kids themselves are expressing unease about AI technology. According to the analysis, a whopping one-third of children in the surveyed countries are worried about how AI is being used. Even scarier, about a quarter fear that their images or videos could be twisted into sexually explicit deepfakes. Yahu, think about that for a second! It’s a wild world out there, and many AI systems hitting the market are doing so with inadequate protections for our children. UNICEF called safety an afterthought, which is just unacceptable.

The agency is pushing for a new approach to global AI governance that embeds children’s rights into the conversation, particularly their right to safety and protection. This is a decisive moment, folks. The choices we make about AI today are going to shape the future for our kids. So, what’s going to happen next? Will we step up and ensure a safer digital landscape for them?

Kaynak: Orijinal Haber

PlayStation 2028’de Fiziksel Diskleri Reddetti: Dijital Çağa Geçiş!

New PlayStation games will no longer be released on discs from January 2028, the gaming giant Sony has announced. In a blog post, they confirmed that

New PlayStation games will no longer be released on discs from January 2028, the gaming giant Sony has announced. In a blog post, they confirmed that while external new games can still be purchased in shops, these will only come with a digital code. This announcement follows closely on the heels of Rockstar’s decision to launch the much-anticipated Grand Theft Auto VI without a physical disc. It’s a major turning point for the gaming industry, which has increasingly leaned on digital distribution in recent years.

Sony’s move is seen as a step towards a fully digital future, but it has sparked significant backlash online. Gaming journalist Vikki Blake has called it a “body blow to consumer rights,” highlighting the implications of this shift for players who prefer physical copies. “It’s not funny anymore, is it?” she remarked, emphasizing the frustration amongst avid gamers.

So, what does this mean for gamers? Well, it looks like the days of collecting physical game discs are numbered. From 2028, if you want the latest titles, you’ll need to embrace the digital realm. This isn’t just a small change; it’s a reflection of the broader trends in tech and entertainment where convenience often outweighs the nostalgia of physical collections.

You know, the gaming community is buzzing with opinions on this. Some folks are excited about the ease of digital downloads, while others are lamenting the loss of a tangible connection to their favorite games. “Vallahi, ben disk koleksiyonu yapmayı seviyorum,” one player commented, showing the mixed feelings around this transition.

Moreover, with this shift, there are questions about how it will affect pricing and accessibility. Will games become more expensive, or will digital sales lead to discounts? And what happens to those who don’t have reliable internet access? There’s a lot to unpack here, and it’s clear that this is just the beginning of a new era in gaming.

Bakalım, bu dijital geçiş sürecinde neler olacak? Gamers are watching closely, and the industry will surely have to adapt to these new consumer expectations…

Kaynak: Orijinal Haber

Finland’s Landline Phones Go Silent: A 150-Year Era Ends!

Finland on Tuesday pulled the plug on analogue landline phone calls after almost 150 years, making it the latest country to step into the digital age

Finland on Tuesday pulled the plug on analogue landline phone calls after almost 150 years, making it the latest country to step into the digital age. This shift comes as part of a global transition towards modern infrastructure, a trend that’s already been embraced by countries like Estonia, the Netherlands, Norway, and Spain. These nations have been rolling out fiber optic cables, which can handle both internet services and voice calls seamlessly.

The last major telecom operator in Finland, which had been maintaining a fixed-line copper wire network, officially marked the end of its service with a nostalgic call involving the firm’s communication and transport agency. During this bittersweet exchange, the two reminisced about their experiences with landline phones. Manner, one of the callers, shared fond memories of his teenage years in London during the 1980s, when he would call home once a week at a set time just to ensure his family was gathered around the phone. Can you imagine that kind of routine today?

They also delved into what the future holds for mobile technologies. The call concluded with a casual “kuulemiin,” which translates to “speak later,” encapsulating the essence of an era that has now faded into history. Interestingly, there were only a few thousand landline-only plans still active, and no new ones have been sold in years. From this point on, only local operators will continue to offer landline services in Finland, catering to a small group of customers who still rely on local calls, according to public broadcaster Yle.

So, what does this mean for the future? As the last echoes of the landline phones fade away, we’re left wondering how this will shape communication in Finland going forward. Is this really the end of an era, or just the beginning of a new one?

Kaynak: Orijinal Haber

E-Scooter İlanları Yasadışı: UK Yollarında Kullanımı Yasaklandı!

Retailers including Amazon, Argos, and Currys are facing scrutiny after an investigation revealed they are advertising e-scooters for use on public

Retailers including Amazon, Argos, and Currys are facing scrutiny after an investigation revealed they are advertising e-scooters for use on public roads and paths, despite it being illegal in the UK. The Press Association PA’s investigation found that both major brands and independent sellers were promoting these vehicles as suitable for road use. However, the law is clear: only rented e-scooters are allowed to be ridden legally on roads, cycle paths, and parks in the UK. Privately-owned e-scooters are strictly limited to private property, and only with the landowner’s permission…

Yani, e-scooter kullanıcıları, yasal olarak nerelerde hareket edebileceklerini bilmeli. Zaten, geçtiğimiz günlerde bir e-scooter ilanı, Advertising Standards Authority ASA tarafından yasaklandığı için dikkat çekti. ASA, bu yasak kararında, e-scooter’ların nerelerde kullanılabileceğine dair yetersiz bilgi verdiği için bu reklamın kaldırılması gerektiğine hükmetti. Yani, reklamda yasal uyarı bulunsa bile, bu durum yeterli olmuyor eğer e-scooter’ların her alanda kullanılabileceği izlenimi veriliyorsa…

Birçok küçük perakendeci hala internet sitelerinde “commuter” e-scooter’lar için ilanlar yayınlamaya devam ediyor. “Şehirde dolaşmak için harika bir yol, trafiğin arasında 15 mph hızla gidin” gibi ifadelerle, bu ürünleri tanıtıyorlar. Diğer bir satıcı ise, “günlük seyahat ihtiyaçları için akıllı bir seçim” olarak tanımlıyor. Yahu, bu durum milletin kafasını iyice karıştırıyor, değil mi? Bakın, bu e-scooter’ların yalnızca kiralamaya açık olduğunu bilmek her kullanıcının hakkı…

Ama işin aslı şu ki, bu tür ilanlar, kullanıcılar arasında yanlış anlama yaratıyor. Çünkü, insanlar bu e-scooter’ları satın alıp, yasal olmayan yerlerde kullanmayı deniyorlar. İşte burada, sürekli olarak uyarılması gereken bir durum var. Yani, yasaların bir gereği olarak, bu ürünlerin yasal kullanım alanlarını belirtmek gerekiyor. Amazon ve benzeri büyük siteler, bu konuda daha dikkatli olmalılar…

Sonuç olarak, e-scooter reklamlarının yasadışı bir şekilde yapılması, hem kullanıcıların hem de yasal otoritelerin dikkatini çekiyor. Bakalım, bu konuda daha ne tür önlemler alınacak?

Kaynak: Orijinal Haber