startup

There is a startup that turns the ashes of your dead into stones. And he is charging more than 2,000 euros for it

by

Cremation is an increasingly common practice after a death and it is not uncommon for many families to keep the ashes or spread in special places. There is a company that has had an idea to turn the remains of our loved ones into a new object: stones. Parting Stone. It is the name of a startup founded in San Francisco that is dedicated to solidifying the ashes of dead people and shaping them into soft, rounded stones. Justin Crowe, the founder, told in a interview with the Wall Street Journal that people usually keep ashes in any closet because it is not something we feel comfortable with, but when transformed into these stones they become a kind of amulet without that negative connotation. “Some families travel with the stones, share them with friends and loved ones, children paint on them; some people keep them in their pockets and even sleep with them under their pillow,” he says. First tableware. that the idea was born after his grandfather’s funeral, a traditional ceremony that made him feel uncomfortable at such a vulnerable moment. After the experience, Crowe began to think about how he could create something new for the grieving process and first came up with an idea: mixing ashes with ceramic glazes. He created tableware with this technique and began to receive orders, the problem was that to make it he only used a small amount of the ashes and he wanted to create something using more. Solidification. The process to turn ashes into stones is quite simple: refine them to a fine powder, mix them with water, shape the stones and cook. Normally about 40-80 stones come out of the ashes of an adult person and they warn that the color depends on each person. The service costs $2,495 for people and $1,195 for cats and dogs, so that’s nothing. The business of ashes. Cremation is a practice increasing around the world. According to Crossing World Group data, This year in Spain, cremation has surpassed burial for the first time, so these types of initiatives are increasingly common. In Spain there is a company called Omneo that turns ashes into a block of wax with an NFC tag. There are also companies that convert ashes in diamondsin trees and even in coral reefs. The most striking proposal of all is that of the Barcelona company Giem Sportswhich proposed the management of spaces within football stadiums to store the ashes of fans. Betis inaugurated a space of this type and they charged 3,000 euros to keep the ashes for a period of 99 years. Image | Pexels In Xataka | More than 300 people have been cryogenized hoping to be resurrected in the future, but no one has proven that this serves any purpose.

A Chinese startup claims to have created its own TPU to compete with NVIDIA. The only problem is that it is three years late

by

A Chinese startup called Zhonghao Xinying (known internationally as CL Tech) has come to the fore with a bold promise. The company claims to have developed an AI chip that not only circumvents Western intellectual property restrictions, but also outperforms NVIDIA’s A100 chip. Which is very good, but also a little bad. Chana arrives. The chip in question has been named “Chana”, and according to SCMP we are dealing with a GPTPU (General Purpose Tensor Processing Unit). Unlike NVIDIA GPUs, aimed at accelerating AI workloads, this is an ASIC, that is, an application-specific integrated circuit designed from the ground up for neural network workloads. promise. According to Zhonghao Xinying Chana, it offers up to 1.5 times the performance of the NVIDIA A100 based on the Ampere architecture. Not only that: it achieves that performance with 30% lower consumption. The startup highlights that the computational cost per unit would therefore be less than half of that offered by the A100 chips. A little history of the company. Behind Zhonghao Xinying is Yanggong Yifan, an engineer formed at Stanford and the University of Michigan. He worked on the development of several generations of Google TPUs and also on the development of Oracle chips, and in 2018 founded this startup in Hangzhou together with Hanxun Zhengan engineer who worked at Samsung for several years. They were joined by other engineers from Microsoft, Oracle, NVIDIA, Amazon and Facebook, they indicate. on Baidu. We are therefore faced with several of those cases of “boomerang talent” with Chinese engineers who are forged in the US and then return to China to create solutions for their own industry. Solutions that do not depend on the West. Yanggong affirms that its chip features “fully self-controlled IP cores, a custom instruction set, and a fully in-house computing platform. Our chips do not rely on foreign technology licenses, ensuring long-term security and sustainability from an architectural perspective.” But. Although the achievement is striking, it is necessary to put it in perspective. The NVIDIA A100 is a 2020 AI GPU, and even with the improvements that this Chinese startup promises, its performance is, for example, far from H100 chips with Hopper architecture that appeared in 2022. Not to mention of the latest Blackwell Ultra chipswhich are currently NVIDIA’s greatest exponent in terms of AI chips. There are also no details about who makes the chip, and one of the candidates it would be SMICwhich has 7nm technology. They are very far away, and they have another problem. The technical achievement of these engineers is certainly notable, but everything indicates that they are still far from what NVIDIA and its competitors are achieving. like AMD or Google with its recent TPU Ironwood. There is another element that works against them: Chinese manufacturers continue without having direct access to the most advanced photolithography on the market, and although it also there is progress from Chinese manufacturers in that sense, competing is certainly complicated without access to the most advanced technologies. Pressure. In 2024 the company achievement revenues of 598 million yuan (73 million euros) with a net profit of 85.9 million yuan, but in the first half of the year the income was only 102 million yuan and had losses of 144 million yuan. The firm has reached an agreement with its investors by which it will have to go public at the end of 2026, or else it will be forced to buy back shares. The financial pressure is therefore notable for the company, which must demonstrate in the coming months that its roadmap is truly competitive. In Xataka | China was no longer supposed to be able to get its hands on NVIDIA’s most advanced chips. Until he found a shortcut in Indonesia

We believed that no open model could outperform GPT-5. A Chinese startup proves us wrong

by

A Chinese startup called Moonshot just launched Kimi K2 Thinkinga gigantic open model with a trillion parameters that has done something that seemed almost impossible: surpass the best proprietary models from companies like OpenAI, Google or Anthropic. If we thought that “Open Source” models could never compete with GPT-5, Gemini 2.5 Pro or Claude, we were wrong. what has happened. This “AI laboratory” had already announced Kimi K2 in July with that gigantic size of one trillion parameters, but now they have released the “Thinking” version with that same size (32 billion active parameters, Mixture of Experts architecture). According to those responsible, the model is capable of maintaining stable use of agentic tools over between 200 and 300 sequential calls. Or what is the same: it can chain long sequences of actions autonomously and apparently without error. The best of all is not that: it is that it surpasses GPT-5 or Claude Sonnet 4.5 in various tests and costs much less than those models. The benchmarks. Those responsible for Moonshot explained how Kimi K2 Thinking achieves the highest scores in Humanity’s Last Exam (general knowledge, 44.9%) and BrowserComp (agent browsers, 60.2%). He is almost at Claude’s level in the SWE software development test, and is also almost the best in another of those benchmarks, LiveCodeBench v6. It is true that in some tests still slightly behind of its “western” rivals, but the achievement is spectacular. More benchmarks. Those responsible for Artificial Analysis have shown their first conclusions after evaluating it with various tests. Thus, they highlight its behavior in agentic tasks that simulate that the model is acting as a customer service agent. In this test it obtained 93% of the maximum, surpassing all its competitors by far (GPT-5 Codex High obtained 87%, for example). They will do more tests, but for now the prospects are fantastic. And on top of that, cheap. On CNBC indicate that training the model cost $4.6 million, a ridiculous figure considering that training proprietary models like GPT-5 It cost about 500 million dollars according to estimates. Using the Kimi K2 Thinking API is also very affordable: $0.6 per million tokens in and $2.5 per million tokens out. GPT-5 Chat costs $1.25/10 respectively, while Claude Sonnet 4.5 costs $3/15 respectively. The details. The model makes use of an INT4 quantization to improve its efficiency without compromising the precision and quality of its responses. Its context window—the “size” of the data we can enter when making prompts—is 256k, a relatively modest figure for large models but still notable. And as a good open model, we can download it to use locally… if we have a real monster at our disposal. The model weighs 594 GB, and for example joining two Mac Studio M3 Ultra It is possible to make it work locally relatively smoothly at about 15 t/s. Alibaba is behindyes. Although the model is developed by an independent startup called Moonshot, this firm has been financially supported by Alibaba, which is becoming an absolute powerhouse in this field. Already not only conforms with developing its own models, which are outstanding (Qwen is the clear example), but is also financing the development of other models such as Kimi K2/Thinking. China and its love for open AI models. During the last few months we have seen how China dominated in the field of open AI models —not “Open Source”—. The Asian giant has adopted an overwhelming philosophy with increasingly better models but which until now seemed to be several steps behind the large proprietary models of OpenAI, Anthropic or Google. This is no longer the case. The race is lively. This achievement represents a new vote of confidence for the open models coming from Chinese companies. It is true that they are huge and that makes it very difficult to use them in practice by end users, but they present an interesting alternative for companies. Image | idnaklss with Midjourney In Xataka | There are many “internal” races within the greater AI race. And Alibaba is winning Open Source

Synthetic influencers are already selling in the thousands. A startup offers them as a service to manipulate networks

by

Influencer accounts created by AI are already a reality, some even have hundreds of thousands of followers. There is a startup that has taken this idea to the next level: they create and manage synthetic influencers to orchestrate massive actions on different platforms, all using AI. Their website reads “Never pay a human again”, a true declaration of intent. Doublespeed. It is the name of the startup that offers the service. Using AI, they create the accounts of these fake influencers and also the content, all with minimal human intervention, just a few finishing touches. Its goal is to “orchestrate actions on thousands of social accounts through the creation and massive deployment of content.” They count in 404media that the startup is financed by the Andreessen Horowitzone of the most important venture capital funds in Silicon Valley. Make it look human. The platforms have systems to detect bots, but at Doublespeed they have the solution to make their AI influencers appear human to the algorithms. In addition, the accounts they offer have been used, since newly created accounts with hardly any interactions are easier to detect as bots. The company’s co-founder, Zuhair Lakhani, said in a podcast that use a “mobile farm” (like the click farms) to manage all their accounts and boasted that one client got almost 5 million views in less than a month with 15 of these AI influencers. Raising the level. He astroturfing It is a tactic through which artificial opinions are generated that seek to appear real and spontaneous, all in order to give an impression of support (or rejection) of a topic or product. What Doublespeed does is next level, creating not only the message, but the “persona” who spreads it. Doublespeed sells “packs” between $1,500 and $7,500, depending on the number of posts they want to generate. Cons the rules. The point is that this practice goes against the rules of the main platforms, such as Meta, which they expressly say that accounts that “make a misleading representation of identity to deceive or confuse people” will not be allowed. It is not the first company to offer services of this type, What is striking is that it has one of the largest funds in the world behind it. dead internet. Is a conspiracy theory which says that the internet is full of bots and humans have been replaced by algorithms. There is some truth in it. According to the cybersecurity company Imperva, in 2024 more than half of internet traffic was non-human. With the emergence of AI, networks were flooded with AI Slop and now it also comes in the form of fake influencers. Image | Reshma Mallecha, Pexels In Xataka | The more we know about the evolution of the internet, the closer we come to a conclusion: bots can kill it

ASML and TSMC are masters of semiconductors. A US startup believes it knows how to end them: with X-rays

by

A mysterious American startup called Substrate has made its appearance with a purpose extraordinarily ambitious: compete head to head with ASML. The Dutch company has become the master of the segment of advanced photolithography machines for chip manufacturing, but at Substrate they believe they have the key to turn the tables. Why is it important. ASML has no competition in the market since it placed its first equipment of UVE photolithography. The ASML thing is a monopoly de facto: If a chip manufacturer or designer wants to access to produce the most advanced models, it depends entirely on the Dutch company. No one has managed to stop it since then, and even China, which is trying to free itself from that dependence, it’s really complicated. Substrate. This is the name of this startup that has developed a new team in which use particle acceleration to manage lithography. This technology allows microscopic circuits to be etched onto silicon wafers, and this new company claims that its machines could be in manufacturing plants in the US within the next two years. It all sounds very good. Maybe too much, but they already have funds to try: they have just raised 100 million dollars and among investors There is Peter Thielco-founder of PayPal and current CEO of Palantir. And already, they will try to help create the new TSMC. The challenge is enormous: ASML has invested decades and billions of dollars to perfect its photolithographic equipment, and the complexity of this market makes it very difficult for companies created from scratch to compete. Substrate’s objective is twofold, because it also aims to ensure that its machines enable the affordable chip manufacturing in the US. Or what is the same: it not only wants to compete directly with ASML, but also allow American manufacturers to compete with TSMC on American soil. There it is nothing. Light is everything. When creating those circuits, some of the lines created are so fine that their dimensions are even thinner than the wavelength of light. To solve that ASML problem they do use of extreme ultraviolet light (EUV) through a very complex set of lenses and mirrors. These machines generate a type of artificial light that is capable of engraving those very fine lines on advanced chips. Very special X-rays. In Substrate they propose a different idea. Although they do not give all the details to prevent someone from copying them, their machines use X-ray lithography (LRX). A particle accelerator creates a light source from x-rays with a shorter wavelength, allowing the beam to be used to create advanced chips. Current Substrate machines are currently capable of working with 12 nanometer nodes, comparable to ASML High NA EUV machines: That would put them on par with the most advanced chip production lines in the world. LRX technology is not new. This technique has existed since the 1970s, but had been abandoned because longer wave techniques (UVP and UVE) continued to scale without the need to overcome the great technical challenges of RXL. In substrate they seem to have precisely the solution to those problems, which focused on optics and the need to use massive particle accelerators as a light source. And they promise brutal cost savings. At Substrate they maintain that their LRX equipment will have an estimated cost of about 40 million dollars, compared to 400 million for ASML’s High-NA EUV. In addition, there would be another associated revolution: the possibility of carrying out the so-called single exposure patterning at advanced resolutions (2 nm, 1 nm and beyond), thus eliminating the additional costs of the multipatterning (multiple exposures). If that promise is fulfilled, the production cost of advanced wafers would be 50% lower according to Substrate. But. Of course, it’s one thing to say it and another to do it. The precision of the process, they explain in Semianalysisis a monumental obstacle. The transfer of patterns (going “from light to silicon” almost “carving it”) remains problematic, and among the challenges is solving the roughness of the edges of the printed patterns, which is amplified during engraving. There is also the problem of stochastic noise (shot noise, random fluctuations in photons that cause defects) and secondary electron blurring inherent to the high energy of X-rays, which fundamentally limits resolution. There are also currently problems with the 1.6 nm overlay, which is still high for more advanced processes where 1.0 or 1.2 nm are sought. In Xataka | AMD is today a semiconductor giant. In its beginnings it was something much more humble: a blatant copy of Intel

The startup that capitalizes our fear of an invisible enemy

by

The places where we can find microplastics grow as time passes. We have already detected them in Testicles, lettuce, at seathe breast milk either even in the air. Taking advantage of this growing concern for this pollutant, there is a London company that is selling an ‘innovative’ therapy: clean the blood of microplastics in exchange for a payment of 11,500 euros the session. It is not yet known if microplastics are in the blood. Today, microplastics have been found in many parts of our body, but the fact that they are circulating in the blood is something that today is not serious for our health. Although companies have seen the opportunity to create a great business around health and this pollutant. A luxury treatment in the heart of London. In the prestigious Harley Street, known for its private clinics and its wealthy clientele, between 10 and 15 people a week they sit on a comfortable armchair to undergo the treatment of the company Clarify Clinics. The procedure, which lasts about two hours, resembles a dialysis session. Although previously we have to pay 9,750 pounds, which would be around 11,500 euros. Something that the big stars pay such as Orlando Bloom. A cannula extract the patient’s blood and introduces it into a machine that separates the plasma from red blood cells. This plasma passes through a filter designed to catch microplastics and other pollutants such as perfluoroalized and polyfluoralized substances. Once the blood is ‘clean’, the plasma is recombine with blood cells and reintroduced into the body. So comfortable that you can work during therapy. In his attempt to attract more and more patients, CEO of Clarify Clinic points out that it is a “very comfortable” technique According to Wired. So comfortable is that he acknowledges that patients make calls, gather by zoom or even watch a movie while their blood is cleaned. Indicated for patients with nonspecific symptoms. According to the company itself, customers who go to their clinic share chronic fatigue or ‘mental fog’. Although the clinic also promotes its treatments for people who take medications to lose weight such as Ozempic, couples seeking to conceive or those concerned about dementia. The result is always the same: they come out as new. Science is not clear. The scientific community has shown that microplastics are literally surrounding us. But there is no evidence of its effects on the body. WHO itselfin a report published in 2019, he concluded that there is still not enough evidence to determine if they suppose a risk to human health. We do not know if they are safe, but we also do not know the risks they could raise to create specific therapies to ‘detoxify ourselves’. If we go to the website of this clinicthe truth is that available from a section called ‘Science’. But upon entering it offers many data, but few references (if not any) of research that speak of the effects on the health of microplastics. They are limited to what we know: their presence in numerous places. But, on the other hand, they point out that they eliminate all the effects of this pollutant. More research is still needed. Although some studies have found worrying correlations, causality remains very elusive. A 2022 review associated microplastics with damage to human cells in the laboratory, but did not examine health results in living people. More recently, In a study Published in the prestigious magazine New England He found that people with microplastics in the fatty plaque of their carotid arteries had a higher risk of heart attack and cardiovascular accident. However, the study was observational and could not prove that microplastics will cause That greatest risk. There could be other factors: patients with microplastics in their arteries were also more likely to be men, smokers and to have pre -existing cardiovascular diseases. A anecdotal -based therapy. Although the evidence is very small, the CEO of Clarify Clinics sells its therapy as a miraculous thanks to the effects it has on its patients. It points to the fact that higher levels of energy or better sleep are reported, and even herself monitors her dream with The oura ring And he assures that after the treatment, his sleep score has not dropped from 90, when before 70 was a good result for her. Biohacking for rich: a booming trend. Microplastic blood cleaning joins a growing list of high -coastal treatments and with doubtful scientific evidence. From the injections of stem cells in Bahamasuntil Receive a child’s plasma injection To stay young. And in many cases these treatments have a common goal: to stay young. And for this there are many people who They can invest millions of dollars in therapies. Our obsession with this objective makes even investigations are pointing to medications that reverse human agingalthough without thinking about the consequences you have. Images | Flyd Ozkan Guner In Xataka | Telling what worries the AI was just the first step: there are already bots that are passed through collegiate psychiatrists

Vlex, the startup that has created “the chatgpt of lawyers”

by

Oakley Capital has sold Vlexthe Spanish legal platform, to the Canadian Clio for more than 1,000 million dollars, according to Expansion. This operation makes the company founded by the FAUS brothers a quarter of a century ago in the sixth active Spanish unicorn. Why is it important. This sale represents the greatest success of Oakley Capital in Spain and confirms the unstoppable boom in the sector LegalTech. The valuation involves multiplying the VLEX value in less than three years, since Oakley bought it for 70 million At the end of 2022. In detail. Vlex operates as the “chatgpt of lawyers”: It offers access to more than 1,000 million legal documents from one hundred countries through its AI ‘assistantVincent ai‘. Its 2.8 million users include from Harvard Law School to Legal Deloitte. Every day incorporates 350,000 new documents to its platform. Oakley Capital, the British manager who in Spain controls idealistic and has invested in Seedgdemonstrated its smell to detect opportunities before boom. When it was done with VLEX control, the company barely generated five million Ebitda. Between the lines. He Timing Oakley has been impeccable: he invested just before the generative AI put the world up, including the legal sector. The pandemia had already accelerated the digitalization of the offices that were still behind, but Chatgpt changed the story. And tools like Vincent ai ceased to be a curiosity, something complementary, to become fundamental. The threat. For its rivals, this operation is a turning point. Cliovalued at 3,000 million and backed by Goldman Sachs, has suddenly been made with the largest digitalized legal library in the world. The consolidation of the sector accelerates. Deepen. The ecosystem LegalTech It continues to attract a lot of money in investments. The best example is Harvey, American competitor from Vincent AI. It was valued at 5,000 million A few weeks ago, when he got heavyweight as investors such as OpenAi or Capital Sequoia. And there is Vlex, demonstrating that Spain is able to compete in the First Technological Division. Not in fundamental models, but in more specialized products, more than niche. Like the new unicorn. In Xataka | Spain is no longer the ugly duckling of the European technological ecosystem. Now has the opposite problem Outstanding image | Vlex, Mockuuups Studio

An AI startup with six months of life and six employees has sold for 80 million dollars. Vibe-Coding, of course

by

Maor Shlomo is 31 years old, is Israeli and six months ago created a small platform of Vibe Coding. He did it almost like a secondary project, but the growth of the project has been vertiginous. So much that after that time the company has just sold for 80 million dollars. We are facing a sign of the times that come to us. AI as a unicorn promoter. There are more “unicorns” than ever. None has become a true giant, but reaching an assessment of 1,000 million dollars has become something relatively normal. Achieving something like that seems very complicated, but there is already talk of how the irruption of AI will make many entrepreneurs convert their startups into unicorns. Uniquersonal unicorns. The difference With the current unicorns It is that these future business successes may be created and managed by a single person. It is at least what it promises according to some AI, which will multiply productivity and avoid having to depend on other people to generate spectacular value. It is already spoken of “Only Unicorns” either “One-Person Unicorns“And the impact that AI agents can have in this type of startups. An example that brings us closer to that future. Shlomo created his little startup, called Base44just six months ago, but at that time the growth of it was such that it ended up hiring six employees According to Ctech. This week he announced that he had sold his company to Wix – a Platform to create blogs and websites – for 80 million dollars. 25 of them which will go to Shlomo and their team as “bonus” to retain them and continue working in Wix – also Israeli – although there are no data for how long they will have to stay in the company to collect said bonus. Vibe Coding. The platform created by Shlomo Perimte users create applications or games without having programming experience. It is an example of That fever for Vibe Coding that we are seeing in the world of programming. In just a few months Base44 managed to attract 100,000 users, in addition to signing various agreements with several Israeli companies known as Etoro or Similarweb. A singular entrepreneur. Shlomo had already co -founded Explorium, a predictive analysis company of Big Data, at age 24. He also made her In a successbut he had to comply with Israeli military service. After completing the service at the end of 2024, he preferred not to return to Explorium and work on another project. He counting The origin and its progress when creating base44 through Your X account. In less than 60 days I already had 100,000 users and two weeks ago indicated that he had generated a benefit of $ 189,000. Far from being a unicorn. Shlomo’s success is remarkable, but of course it is far from being considered a unicorn and much less One of those “only unicorns” “It has several employees in your team,” that is spoken so much and that the theoretically promotes. And yet, it is a good example that artificial intelligence raises the future. The AI ​​agents are in diapers, but the promise is that they will automate a lot of processes for those who use them. That could impact significantly on the template that these future startups will need, but for now everything is, we insist, a promise. One that Shlomo has become a reality. Image | Christina In Xataka | India has its own ‘Silicon Valley’ in Bangladés. The problem is that it is a ghost city

A Basque startup of AI has just lifted 189 million euros with a great idea: compress the AI

by

Before We compressed files with Zip. Now what we begin to need is to compress the AI ​​to make it more small and efficient. That is just the idea that the founders of Multivrse Computing had, a Spanish startup which is becoming the new jewel of the crown of our AI industry. Its founders, (in the image, from left to right, Román Orús, Enrique Lizaso Olmos and Samuel Mugel) and Alfonso Rubio have much to celebrate. Investment Round. Multivize Computing He has just closed an investment round of 189 million euros (215 million dollars). The round (series B) has been led by Bullhound Capital, but it has also participated HP Tech Ventures, Sett, ForgePoint Capital International, CDP Venture Capital, Santander Climate VC, Quantonation, Toshiba and Euskadi Risk Capital of Euskadi – Spri Group. Last March the company received An investment of 67 million euros by the Government of Spain. The inference AI by flag. Although the current prominence usually takes it the great technological ones that invest billions of dollars in data centers to train Great language models (LLM)there is more and more focus on the other part: the one we use users when asking things to Chatgpt, for example. It is the so -called AI inference, and the estimate is that in 2025 the value of that industry reaches 106,000 million dollars. In Multivrse Computing they want a good piece of that cake, and to achieve this, its great trick is a unique technology. Compactifai. This is the name of The compression technology of AI models developed by multivance computing. What this allows is to convert very large models – which costs a lot to “execute” – in much smaller and efficient models, which allows them to make them more manageable and save many resources (and time) during inference. How to compress an AI model. Román Orús, scientific director of the company, led A study May 2024 in which they precisely explained the concept of “tensioning networks” of quantum inspiration and that allow compressing these models. Its operation is based on decomposing the matrices of pesos from the neural networks “truncating them” and retaining only the largest and most relevant values. In essence the concept focuses on discarding the less relevant information of the model to be left alone with the most relevant. But that does not make the model less accurate? In fact, but the degree of truncation can be controlled so that there is a good balance and commitment between compression and loss of precision. Even by compressing these models, in Multivars Computing they say The fall of the models It is only 2 to 3%. Same yield in a size 95% lower. To mitigate that precision fall, this system includes a rapid resentment phase called “healing” that can be repeated several times to achieve even closer accuracy to the original version. In the end, they affirm in the company, they can compress up to 95% a model of the performance. It lowers the use of AI. According to Your dataa model as it calls 3.1 405b has an operational cost of about $ 390,000 if we want to run it at home (13 GPUS H100, 9100 W of consumption), but thanks to Compactifai it is possible to reduce that cost to 60,000 dollars (2 GPUS H100, 1,400 W). One more “thin”. The “Slim” models provided by the company – Derivatives of Llama 3.3 70b or Call 4 scout– They are compressed versions that theoretically do not lose precision. They can be executed through the AWS platform or by licenses that also allow us to use it on-premisethat is, in local/own infrastructure. According to their metrics, these models are between 4 and 12 times faster than their non -compressed versions, which translates into an inference cost that is between 50% and 80% lower. Image | Multivize Computing In Xataka | Spain is finally

Using a mobile is a real challenge for older people. An Aragonese startup has a promising solution

by

Is called Maximilianabut it is not a person. It is a mobile. Specifically, one specifically designed for our elders. The idea that this device raises is unique, already a difference from other mobiles for the elderly, everything is based on trying that these people have to touch anything to communicate with their loved ones both in video calls and normal calls and to alert possible emergencies. Innovating from Aragon. Many business projects are born from personal needs. This happened for Jorge Terreu (Zaragoza, 26 years old), who a few years ago was Erasmus in France and had trouble contacting his grandmother, Maximiliana. This is how the germ of her project was created: create a phone to be able to talk to her without her having to do anything. @Miguessal This mobile can save your grandfather’s life, because it makes video calls down alone. It was created by a grandson for his grandmother, Maximiliana, who did not manage with technology. What began as a gift, is today a mobile that thousands of people use. You can learn more about him at www.maximiliana.es. #Salvarvidas #Emercances #Primerosauxilios #Urgences #security ♬ Original sound – Miguel Assal The principle of something big. Terreu prepared a mobile prototype in which he started the video call and the phone answered alone. He sent it to his grandmother in Madrid and soon realized that this idea could serve many more people. In March 2020, during the pandemic he donated five phones adapted to a hospital in Zaragoza and verified how users were delighted, so he launched the adventure of creating his own company and his project of a mobile for seniors. What would be called, of course, like his grandmother. Maximiliana. They hardly have to touch it. People who use Maximilian mobiles hardly have to touch it, because they can receive video calls or messages without having to play in the interface. If a relative or friend makes a video call, Maximiliana sounds for a few seconds and automatically breaks. If it is the elderly the one who wants to call, the interface is especially intuitive and simple and is based on a “desk” on which the faces of the family appear. Just press one of them to make the call. Emergencies. Maximiliana also has emergency call options. Thus, if for example the elderly falls and cannot handle the mobile with precision, it is enough that it agitates it a bit to activate the emergency call to the assigned contact. In those calls the mobile uses the flashlight to emit the international relief signal. Always located. Another of the Maximiliana options It is what allows to know at all times the location of that elderly, something especially useful if they are lost or misplaced. The software also shows the battery available so that family members can warn that it loads it to continue using it without problems. THE MOBILE AS SERVICE. Maximiliana includes the mobile, the control app for the family and the accessories, such as the cord or antiahogo strap that allows the device to be hung by the neck. The product is not sold as a mobile without more, but as part of a subscription service. The “basic pack” cited costs 24.90 euros per month, but it is also possible to directly hire the SIM card (unlimited calls, 25 GB of the Internet) as part of the “complete pack” for 29.90 euros per month. Includes apps like WhatsApp. Maximilian mobiles remain mobile, and allow to continue accessing applications such as WhatsAppChamber or Gallery. The person in charge of the device – normally, a relative of the elderly – chooses from the control application which applications wants them to appear on the mobile. It is the relative who acts as “administrator”. The relatives of the people who use Maximiliana are the ones in charge of managing and configuring the mobile options. Thus, they can add contact faces, change order (up to six on the screen), or adjust the sound duration. A grandmother like Community Manager. As they pointed out In the countryMaximiliana’s team has been growing as the success of the service has been curdling, but there is a very special component in that template. Your Community Manager It is Conchita Polo, grandmother of Pedro Malo, partner of Terreu in the project. More than the phone matters the service. There are currently good options when offering Mobile for seniors And also tips for Adapt any Android phone To make it easier to use for them, but Maximiliana’s proposal is interesting because everything combines. Take advantage of the idea of parental controls that for example allow to know the location of our children or control the limits of use, but adapting it to a population sector that can certainly take advantage of it in a beneficial way. Image | Maximiliana In Xataka | What five mobiles would I recommend to give my grandfather this Christmas, for simplicity and quality-quality

Log In

Forgot password?

Forgot password?

Enter your account data and we will send you a link to reset your password.

Your password reset link appears to be invalid or expired.

Log in

Privacy Policy

Add to Collection

No Collections

Here you'll find all collections you've created before.