advance Archives - Page 3 of 3

GPT-4.5 It is not better than its rivals in almost anything. It is the proof that traditional AI models almost do not advance

February 28, 2025 by usatoday24

Sam Altman I had already warned that they planned to launch GPT-4.5 very soon. We had been waiting for the GPT-4 successor for months, but over time expectations have been going down: there was talk of the Founder of AI And how climbing – more data and more GPUS to train models— It didn’t work so much. Precisely GPT-4.5 was going to be proof that perhaps that was not true. Do you know what? That was probably, because GPT-4.5 is a model with many starting problems. GPT-4.5 is already with us. Yesterday Openai finally presented GPT-4.5the theoretical successor of GPT-4. Sam Altman explained that this was “the first model that makes me feel that I am talking to an attentive person.” Gigantic and expensive. But Altman also recognized something else. “Bad news: it is a giant and expensive model.” The head of OpenAI claimed to have run out of sufficient GPUS to make a mass launch, and the availability of GPT-4.5 is very limited: only Chatgpt Pro users can use it for the moment. Caro no, very expensive. Using the GPT-4.5 model through the OpenAi API is extraordinarily expensive: it costs $ 75 per million input tokens, and $ 150 per million output tokens. GPT-4O costs 2.5 and 10 dollars respectively (30 and 15 times less), and O1, so far the most expensive, costs 15 and 60 dollars respectively. And it is also not a “border” model. He Technical Report OpenAi indicates that this It is not a model “Frontier“ As was GPT-4, for example. That is important, because despite being its largest LLM, the border models are more capable, of large scale and raise risks to generate misinformation or be forced to get out of the standards. In GPT-4.5 they seem to have focused a lot on avoiding errors (it is one of its advantages, it seems to put less the leg according to some test banks). It does not seem better in almost anything. The evidence and benchmarks to which it has been subjected seems to make it clear that the leap into benefits is especially disappointing, especially if we compare it with the new models of its rivals. Is worse in accuracy of the facts that perplexity deep researc, is worse than Claude 3.7 Sonnet in programming According to TechCrunch and Several expertsand it is also worse in reasoning (although it is certainly not oriented to it) than Deepseek R1, O3-mini or Claude 3.7 Sonnet (which is a “hybrid” model). Bittersweet feeling. Experts like Simon Willison either Andrej Karpathy They have shared their first impressions and in both cases the sensation is that GPT-4.5 It is slowis updated only Until October 2023 And it does not represent a really remarkable advance. Willinson came to analyze the debate that dozens of users maintained on GPT-4.5, and in a Summary generated by AI The conclusions were also clear: the numbering itself was inappropriate, the model is too expensive, the price/benefits ratio was very debatable and the performance was not what was expected after so much time. Karpathy’s conclusion is that “it is a little better and that is great, but not exactly in trivial sections of highlighting.” More human? Altman’s appreciation about his conversation how he had been surprised by the conversation capacity of GPT-4.5 Maybe he points to the direction in which this model stands out. Karpathy also pointed to that aspect in saying that the improvement could be shown in “creativity, realization of analogies, general understanding and humor”, which perhaps makes effects effectively with GPT-4.5 give the feeling of being even closer to those we would have with a human being. The climb does not work, the deceleration is here. GPT-4.5 It is a clear example of how we have reached the limits of the scaling. Having a gigantic LLM no longer seems to provide advantages over its predecessors, and dedicate more data and more GPUS to train these models does not seem to make much sense. Altman himself made it clear that GPT-4.5 would be the latest non-reasoning model of the company. That is another sign that demonstrates that the deceleration of the generative AI, at least in regard to traditional models, is a reality. Why have you launched it then? In it OpenAi blog It indicates how “we are sharing GPT-4.5 as a research advance to better understand its strengths and limitations. We are still exploring what they are capable of and we are eager to see how people use it in ways that we would not have expected.” That seems to show doubts that their own creators have with the model, and the question why they have thrown it. They need to continue generating “Hype “. Especially considering that the rivals are very strong lately. Claude 3.7, Grok 3 and of course Deepseek R1 have managed to turn the tortilla and raise a challenge for Openai, which until not long ago seemed to be a step ahead of their rivals. Now that is not clear, and in many sections its competitors already exceed the benefits of their models. OpenAi needs to breastfeed and say “here I am”, but perhaps with GPT-4.5 that movement goes wrong, because at least a priori the benefits are disappointing. And investors squeeze. Some point to another probable theory for this launch. OpenAi could have been forced to launch GPT-4.5 make investorsthat have invested billions of dollars in the company and that need to be calm with their investment. Once again OpenAi has a problem, because it does not seem that GPT-4.5 can leave them calm. It will be difficult for new investors to be convenient with this launch. In Xataka | Openai has a golden opportunity to sweep all its rivals: launch an unlimited chatgpt and full of advertising

I have run for weeks with a Zonense Suunto. It is a spectacular advance for both Marathonians and those who just want to lose weight

February 28, 2025 by usatoday24

In recent weeks I have run a few times with a Suunto Race On my wrist next to a pectoral heart rate. Was necessary to try Zonense, the novelty presented a few months ago by Suunto. I wanted to understand her well and try her for myself. The truth is that it is an inventazo. What exactly Zonese is? Zonese is Suunto’s new technology to measure our intensity during exercise. The usual thing is to count pulsations per minute, but Zonese analyzes the Heard rate variability (VFC) To determine, in real time, how our body is responding to effort. The VFC is the variation in time between each heartbeat. Not all beats have exactly the same temporal separation, and these small variations contain valuable information on how our autonomic nervous system is working. High variability usually indicates a good state of recovery, while a low variability indicates fatigue or stress. The interesting thing is that This VFC changes during the exercise very specificly according to the intensity with which we work. And it is just this relationship that Zonese takes advantage of an algorithm called DDFA (Analysis of dynamic fluctuations without a trend) to determine in which metabolic area we are training. And that serves to optimize anyone’s training. Either an expert athlete who wants to increase his performance, or either a person who simply wants to go for a walk to lose weight and wants to do it in the optimal way. Three areas that change everything The training areas, or heart rate, are five, although according to whom you ask (including sports watches manufacturers) can give you six or seven. Zonense simplifies training only three colors: Image: Xataka. Green (aerobic zone). When your body is working mainly with oxygen and fats as fuel. It is the ideal zone for long training, recovery or fat burning. Your body could be kept here for hours. Do you want to go for a walk to lose weight? This is your area. Do not stay below or above. Yellow (anaerobic). Your body begins to use more glycogen and begins to produce lactate, although it can still process it. It is the “threshold” area, where performance improvements are important, but you can’t stay here for too long. Red (Vo2max). You are above your anaerobic threshold, your body produces lactate faster than you can eliminate. It is where the greatest adaptations of power and speed are achieved, but you can only maintain this effort during short time spaces. He Vo2max It is the maximum amount of oxygen that your body can process during intense exercise, is what measures our aerobic capacity. Why Zonese needs a heart rate A key aspect of Zonese is that it requires a heart rate. The optical wrist sensor that incorporates the watches is not enough, and there is a scientific reason for it. Optical sensors They measure the pulses by calculating changes in blood flow by lightbut They cannot accurately detect the exact intervals between beats nor the small variations necessary for the VFC analysis during the movement. Chest bands, however, directly detect the electrical impulses of the heart (such as an ECG), providing the necessary precision for this advanced analysis. In my tests I have used the Suunto Smart Sensor. It would be much better not to have to use it and limit us to the clock, but if these bands maintain their popularity it is for something: they are more reliable, precise and complete. My experience The first time I went to run with Zonense, I planned a 10 km training at a soft rhythm. Normally, based on my heart rate, I would run at about 5:30 min/km to keep my “aerobic zone”. However, after the initial 10 minutes that Zonese needs to calibrate (establishing your metabolic baseline of the day), the clock showed me that it was in yellow (anaerobic), despite the fact that my heart rate was just 145 ppm, which would traditionally consider as my area 2-3. It turns out that last night (several, in fact) he had slept badly and also came from three consecutive days of intense training. My body was telling me, through Zonesethat although the pulsations were not extremely high, it was metabolically working harder than traditional numbers suggested. I reduced the rhythm at 5:47 min/km and, indeed, I returned to the green zone. The training was much more pleasant and, surprisingly, the next day I felt quite recovered, something that would probably have happened if I had maintained my “normal” rhythm. On another occasion, after a week of rest, Zonese allowed me to maintain a faster rate (5:15 min/km) remaining in the green zone. My body was cooler and could work at greater intensity without entering anaerobic areas. Image: Xataka. This type of Personalized and dynamic adjustment is what makes Zonense so valuable. They are not just numbers on a screen, it is real information on how your body is responding to exercise at that precise moment. Another striking example: the shoes. With which I used for filming I marked VO2Max from 4:47 min/km, while when I went to flying I had to go down to 4:19 min/km to enter that area. Not only for elite athletes One of the biggest surprises has been to verify the usefulness of Zonese for people who are barely starting to play sports or simply go for a walk. It is common for a beginner to feel overwhelmed at the explanations of cardiac areas and to remove calculations from its theoretical maximum frequency. For this type of users, Zonese says something much simpler: “Stay in green”. With that simple instruction, the user will walk at a pace that maximizes fat burning without entering areas of greater intensity where the body begins to use glycogen. And it will avoid going too slow. And best of all: being real -time and calibrated information after ten minutes of activity, that user may be in the green or yellow zone depending on the day even if … Read more

convert the escape of Helio de la Starliner into a great military advance

February 25, 2025 by usatoday24

A few weeks ago, the two “stranded” astronauts in space after leaving for the International Space Station on June 5 aboard the Starliner ship, received good news: He had decided to advance his return. For Boeing, however, the mission fiasco is far from finishing. In fact, the company is Trying to sell your bleeding space division by piecesl. The latest: it seems that China has taken “oil” from the main problem of the Starliner. Boeing’s problem as inspiration. In a surprising return of technological nut, Chinese scientists seem to have transformed the serious crisis that affects NASA (and especially Boeing) in A revolution in aerospace propulsion. The Boeing Starliner capsule suffered multiple helium leaks that incapacitated their propulsion systems, leaving the two astronauts trapped at the International Space Station. That incident turned the heliuman ultralight gas traditionally used to pressurize liquid fuels, in A symbol of technical vulnerability. Now, Chinese researchers have seen in that same weakness a unique opportunity to develop a disruptive technology, a Able to revolutionize both the military and spatial field. Solid propulsion driven by Helio. The team of scientists, led by Yang Zenan of the Harbin University of Engineering, has revealed in a recent study Published in the Magazine Aeronautica et Astronautics Sinica How helium controlled injection can considerably multiply the performance of solid fuel rocket. Specifically, they discovered that by introducing helium through microscopic pores (about 2 mm) in the combustion chamber, It is possible to increase the specific engine efficiency by 5.77%. In addition, by precisely adjusting the proportion between helium and combustion gases (a part of helium for every four parts of fuel gas), scientists They managed to instantly triple the engine thrustcarrying it from 100% to 313% according to operational needs. Advantage: Thermal camouflage and invisibility. The use of helium is not limited to potential improvements or thrust. One of its most strategic advantages lies in the drastic reduction in the temperature of the exhaust gases. The helium, Being extremely light (With a molar mass of only 4 g/mol in front of the 29 g/mol of the typical combustion gases), expand and cool the escape flow, lowering its temperature by up to 1,327 degrees Celsius. This thermal reduction implies a radical decrease in the infrared signal of the missiles propelled by this technology, making them almost impossible to detect through infrared sensors modern military satellite employees such as Spacex Starshield and interceptor systems such as The SM-3 Block IIA. Also improved stability and safety. Another relevant feature of helium, highlighted by researchers, is Its chemically inert naturewhich avoids common problems associated with unstable combustion caused by other light gases, especially hydrogen. This ensures, they explain, that The flight is soft and predictableeliminating many operational risks that traditionally hinder the practical application of more volatile alternative technologies. In addition, by absorbing excess heat generated by combustion, helium not only reduces the exhaust temperature, but also Protects internal engine materials, extending its useful life and decreasing operational risks. Military and strategic applications. They count on work that from a strategic and military perspective, technology has deep implications. The Ability to quickly modulate the thrust of a missile in flight allows Unpredictable attack trajectoriessignificantly improving your maneuverability. This aspect is especially relevant to advanced defensive systems, since the combination of lower infrared detectability and variable acceleration capacity makes interception greatly difficult, especially in the context of defense against hypersonic missiles. According to Chinese researchers, this innovation could directly solve the current challenges facing their weapons systems in terms of speed, scope and thermal stealth, presenting a real challenge to antimile defenses developed by the United States. Civil and spatial potential applications. However, the benefits of this technological innovation transcend the military. According to the research team, this technique could also revolutionize the civil and commercial field, especially in The rapid and economical launch of satellites by solid fuel rockets. As? The capacity would be particularly valuable for urgent deployments in commercial, scientific or humanitarian contexts, providing an agile, profitable and adaptable solution against the expensive current launch systems. Current context and relevance. Meanwhile, in the United States the initial crisis continues. As we explained at the beginning, NASA recently announced that astronauts Butch Wilmore and Suni Williams, affected by the Starliner’s breakdown, Finally they will return to Earth in mid -March using a Spacex ship. In the words of the NASA Traffic Program manager, Steve Stich, the incident has demonstrated once again that “the human space flight is full of unexpected challenges.” Paradoxically, that same challenge that affected NASA and Boeing has served for Chinese scientists to achieve, on paper, a disruptive innovation that could deeply redefine the rules of technological and strategic game in the near future. Image | POT, Gaystock In Xataka | After the fiasco of the Starliner ship, Boeing is trying to sell its bleeding spatial division by pieces In Xataka | The two “stranded” astronauts in space have received their first good news in months: their return will be advanced

If it is a matter of time for AI to enter our cars, Nissan and Deepseek have decided to advance

February 14, 2025 by usatoday24

Deepseek’s adoption is being much faster than that of Chatgpt. The artificial intelligence that arises as the new reference in China is already integrating natively into smartphones manufacturers, such as Honor and Nubia In your native country. He is also landing in electric cars, such as ByD, which have already signed agreements to improve their autonomous driving system thanks to Deepseek technology. Following his steps, the first Japanese manufacturer to bet on this Open Source model for greater interface level integration has been Nissan. A plan similar to the byd. This week we knew byd’s plans to integrate Depseek into its research and development line. The agreement with Byd had as its main focus to implement an advanced autonomous driving system, raised for both high -end vehicles and for the most accessible. Beyond the autonomous driving plans, improvements were raised in the vehicle’s own software thanks to the integration of this AI. Is exactly what he wanted Announce Nissan. The company has revealed that its new electric sedan, the Nissan N7 (a vehicle born of a joint venture with the Dongfeng and oriented to the Asian market), is the first vehicle to incorporate a deep integration with Depseek-R1. The vehicle will be able to process natural language, make recommendations related to navigation and entertainment system, as well as adjust to driver’s habits to vary their behavior. A thrust for software. From a time to this part, the software has become one of the main keys when it comes to a vehicle. Android Auto, Android Automotive, Apple CarPlay, Own Solutions of the manufacturer, Alliances with video game development companies… Who wants to lead the car of the future has to lead in software, and alliances with agents such as Depseek will be key to it. China is also one of the countries that is leading the transition to the electric vehicle, and having the main alternative to Chatgpt can completely shake the market. Nissan is not alone. Nissan and Byd are not the only companies that have approached Depseek. The gigantic Chinese group Geely, owner among others of Polestar and Volvo, has also opted for China to integrate it in the future in some of its vehicles. The objective and advantage for the user? The same as in the previous case: that drivers can enjoy a more connected driving experience, thanks to the natural understanding of language that these models allow. Image | Nissan In Xataka |The brief love story between Nissan and Honda is over. Now another buyer appears on the horizon: Foxconn

Champions League: results that Real Madrid needs to advance among the top 8

January 22, 2025 by usatoday24

In addition to a miracle… and that not all those at the top win (between fifth and fifteenth position), The odds for Real Madrid to qualify for the Champions League, among the top 8 in the table, are very tight.. The first favors began on Tuesday, with the epic victory of the Barcelona facing Benfica (5-4), the defeat of Bayer Leverkusen in view of Atlético de Madrid (2-1) and especially that of Liverpool against Lille (2-1). Because they are direct rivals when it comes to counting points, these results increased the odds for the team he leads. Carlo Ancelotti. In fact, the tie Juventus and the defeat of Borussia Dortmund against Bologna they also serve Real Madrid. But there are many games ahead and The first requirement that the Spanish team must meet is to beat Salzburg this Wednesday and Brest next week. If you do not win both duels, you will not be able to qualify among the top eight. The results that Real Madrid needs to advance to the 8th round of the Champions League directly This Wednesday 9 matches will be played and Real Madrid, from the Santiago Bernabéu, will have to pay attention to other duels. The most important will be Feyenoord – Bayern Münich, Leipzig – Sporting Lisbon, Milan – Girona, Sparta Prague – Inter Milan and Shakhtar – Brest. Real Madrid in 22nd place, but If they beat Salzburg and some results merge, they could reach sixteenth place, leaving the rest of the aspirations for next week. Real Madrid needs these results today: Shakhtar’s victory against Brest Leipzig victory against Sporting or draw Sparta Prague victory against Inter or draw Bayern Münich’s victory against Feyenoord Girona victory against Milan or draw If the forecast is fulfilled, according to an online calculator that the Marca newspaper has availableReal Madrid would reach 16th place with 12 points, while Atalanta, who beat Sturm Graz 5-0, is in fifth position with 14 points. And what results does Real Madrid need on the last date of the Champions League group? Basically, defeat Brest and not let everyone at the top win, especially: Atalanta, Bayer Leverkusen, Aston Villa, Inter Milan, Monaco, Lille, Borussia Dortmund and Juventus. However, some of these clubs play at home and against modest rivals. Assuming that Leverkusen wins at home against Sparta Prague and Barcelona defeats Atalanta in Montjuic, perhaps the “easiest” matches to predict, in the rest of the matches ANYTHING can happen. The results that would classify Real Madrid on the last date: Victory or draw for Benfica against Juventus Draw between Inter Milan and Monaco (the victory of either would hurt them) Victory or draw for Feyenoord against Lille Victory or draw for Girona against Arsenal (unlikely) Win or draw for Dinamo Zagreb against Milan (unlikely, but the Croatian club is local) Victory or draw for Shakhtar against Dortmund Barcelona victory over Atalanta But the margin of error is VERY small. If Juventus wins and there is a winner between Inter and Monaco, Real Madrid could only be ninth, unless Arsenal and Bayer Leverkusen lose. And in the event that Arsenal, Bayer Leverkusen win and there is a winner between Inter and Monaco, but Juventus falls or draws against Benfica, Real Madrid’s classification could depend on Barcelona defeating Atalanta. The number of goals scored and conceded, which results in the goal difference, could also be an ultimately determining factor, but difficult to predict at this time. Real Madrid has it complicated, very complicated, but mathematics keeps it alive. Why Real Madrid seeks to be among the top 8 With this new format in the Champions League, the top eight teams advance directly to the round of 16, an objective that Real Madrid seeks in this initial phase. In case of finishing between 9th and 24th place, the whites would have to play a kind of direct eliminatory, round trip, to decide the others classified for the round of 16, where the competition continues with the usual format. If they fall in this instance, there will be no consolation ticket to the Europa League, but rather they will be definitively out of European competitions.

Log In

Sign In

Forgot password?

Your password reset link appears to be invalid or expired.

Log in

Privacy Policy

Add to Collection

No Collections