Kimi K3 triggers the alarm in Silicon Valley

Kimi K3 has just dismantled an idea that had been installed among analysts for many months: the Chinese laboratories of artificial intelligence (IA) were, at least, between six and eight months behind its American rivals. This model, published last week by Beijing startup Moonshot AI, has 2.8 trillion parameters and has been designed for programming, knowledge work and complex reasoning tasks, as well as incorporating native vision capabilities. In fact, it’s bigger than the latest flagship models launched by Chinese rivals, such as Zhipu AI or DeepSeek. Its performance, and this is what is really surprising, has surpassed that of the main closed models from the USAsuch as Claude Opus 4.8 from Anthropic or GPT-5.5 from OpenAI, leaving only from behind of the most recent Claude Fable 5 and GPT-5.6 Sol, according to Artificial Analysis data. China is only a few weeks behind The most shocking conclusion Ryan Fedasiuk picked it upresearcher at the American Business Institute, in a report published this Saturday under the title “China has caught up with the US in frontier AI.” Fedasiuk maintains that the gap between the two countries, which until now was measured in months, has narrowed to something close to a few weeks. Large American laboratories continue to lead the absolute frontier with Claude Fable 5 and GPT-5.6 Sol And Kimi K3 not only competes in performance. In tests such as the Arena AI ranking for development front-endthis Chinese model has surpassed both GPT-5.6 Sol and Claude Fable 5 despite being noticeably cheaper than both American alternatives. Of course, its price is somewhat higher than other Chinese models, such as DeepSeek or those from Zhipu AI. This combination (a larger, more capable and cheaper model than much of the American competition) is what has triggered concern in Silicon Valley, a location accustomed until now to taking for granted a cushion of several months of advantage. In any case, the large American laboratories continue to lead the absolute frontier with Claude Fable 5 and GPT-5.6 Sol. What has changed, according to the Fedasiuk report, is the speed at which that distance is shortened. Silicon Valley can no longer continue discounting that advantage for months. Now it is measured in weeks. Image | Generated by Xataka with ChatGPT More information | SCMP In Xataka | China has a plan to win the AI ​​war against the US. And DeepSeek is its champion

Peter Thiel is one of the pillars of Silicon Valley. His record-breaking move to Florida is one more nail in California’s coffin

A rental contract does not usually provide enough space for owners, except if it is for a centrally located apartment in Madrid or Barcelona for less than 600 euros. Then, it would be a miracle. But when the signer is one of the names that they defined Silicon Valley and their destination is Florida, things change. As and how I collected the North American Fox BusinessPeter Thiel has just signed the contract for what will be his offices in Miami, paying more than anyone else in that city before. “For Peter Thiel to sign that lease, marking a milestone of $250 per square foot, is absolutely incredible,” said Joe DaGrosa, founder and president of the real estate investment consultancy. DaGrosa Capital Partners. The record price reached in office rentals draws attention, but what is truly interesting is what lies behind it: a historic name in Silicon Valley is willing to pay more just to leave California as soon as possible, joining (and urgently) the migratory wave of millionaires who have already established their residence in front of the warm waters from Miami. A record rental that hides a trend Thiel’s office will occupy part of the 44th floor of the tower 830 Brickella busy building in which it will share elevators with companies such as Microsoft or Citadel. Its offices will occupy about 1,690 m2 and the price that has registered the record is $2,690.98 per square meter. That is, a bill of 4.55 million dollars per year. As Joe DaGrosa explained to Fox Businessthe average rent in that area was around $40 or $60 per square foot for an office, which implies that Peter Thiel would be paying between 525% and 316% surcharge for the rent of their offices. The increase in the price of business land in South Florida already compete with the prices of Manhattan and San Francisco. This price increase has arisen from the growing interest of technology companies that until now had their headquarters in Silicon Valley, but are now moving en masse to Miami, following in the footsteps of their founders. Thiel hasn’t arrived in Florida overnight. Just like other millionaires have done like Jeff Bezos o Ken Griffin, founder and CEO of Citadel, Thiel bought a property in Miami Beach more than five years ago. In December 2025, with California already debating the tax on great fortunes, Thiel Capital announced that opened office in that same area. The final blow came in February with the announcement of Palantir: the company chaired by Thiel confirmed the move from its headquarters to Miami, thus completing the migration of Thiel’s main businesses to Miami. Expelled by a tax that does not exist Behind so much movement there is a very specific reason: tax payment. California has been debating for months application of a tax single 5% on the assets of those who exceed $1 billion and reside in the state on January 1, 2026. This tax is expected to raise about $100 billion over the next five years, which will be used to mitigate cuts in social benefits for the most vulnerable population in California. The measure has already gathered more than 1.5 million signatures, well above the million necessary, and was confirmed to be submitted to vote next November after the failure of negotiations between the driving unions and Governor Gavin Newsom, who is against imposing the new rate that would affect about 200 people. That is, the millionaires already have started their preventive migration reacting to a tax that has not even been approved and that does not even have the support of the state governor. Larry Page and Sergey Brin left California before the tax deadline. Page paid more than 170 million dollars for two houses in Coconut Grove, and Brin was bought a 51 million home on Allison Island. Mark Zuckerberg, after two decades in San Francisco, changed zip code to become Jeff Bezos’ neighbor in Billionaire Bunker. The great beneficiary of the move It’s been Florida. His choice has not been causal, and responds to a compelling reason: he does not charge income or wealth tax, and that weighs a lot when the fortune is counted in billions of dollars. Not all technology magnates take a dim view of the tax on large fortunes. Jensen Huang, CEO of NVIDIA, has expressed his willingness to pay taxes get approved and stay in California. But Thiel already donated three million dollars to the California Business Roundtable, the group fighting the proposal. In Xataka | More and more millionaires and technology companies are fleeing California. The winners: Texas and Miami In Xataka | Billionaires prepare their mansions for the apocalypse: luxury bunkers have become a trend Image | Gage SkidmoreUnsplash (mark)

The IMEC laboratory has found a way to extend the life of silicon

IMEC just updated your semiconductor journey and the fact that attracts the most attention is a date: 2038. That is the year in which, according to this Belgian research center, the industry will begin the production of class 3 angstrom integrated circuits (0.3 nanometers). This is not the first time we have talked about this milestone; a little over a year ago We analyze in Xataka a slide from IMEC that placed this leap in 2035. The new forecast delays it by three years, but in exchange it gives us something much more valuable: how to get there. And to reach 0.3 nm it is not enough to improve photolithography alone. IMEC states that the contact poly pitchwhich is the minimum distance between transistors that for decades has been the main indicator of technological progress, will no longer be significantly reduced starting with the A10 generation, which is scheduled for 2030 or 2031. From then on, to increase density it will no longer be enough to shrink the transistors; They will have to be stacked. This paradigm shift has its own name: CFET transistors (Complementary FET). This strategy is not new, although until now it was a distant promise. And the interesting thing is that the IMEC itinerary gives it a date, context, and connects directly with everything we have explained in our articles dedicated to IT teams. UVE Hyper-NA photolithography. These machines will be needed to manufacture these chips, although they are still being developed by ASML. Why CFET technology is going to retire GAA transistors The transistors Gate-All-Around (GAA), which the semiconductor industry began to adopt massively in the 2nm generationthey still have a journey. IMEC estimates that this architecture will remain viable until the A10 generation, which will arrive in 2030 or 2031, giving it a useful life of about seven years from its introduction. It is a reasonable period of time if we compare it with previous generations of transistors, but the Belgian research center already makes it clear that it has an expiration date. The underlying problem is geometric. The GAA transistors improved the electrostatic control of the channel by completely surrounding it with the gate, but they still placed the type materials n and p one next to the other, in the horizontal plane. This provision has an obvious physical limit: There will come a time when it will no longer be possible to continue reducing the distance between them without compromising the electrical performance of the chip. This is precisely what will begin to happen when the A10 generation arrives, according to IMEC. Sequential CFET technology will come first, followed by linked CFET structures CFET transistors solve this problem by stacking the type material n directly above type pvertically. IMEC’s ​​itinerary places their arrival as firm candidates for chip production in the A7 generation, which is scheduled for 2033. And, in addition, it explicitly associates them with the need to resort to power delivery systems on the back side of the wafer, which IMEC considers mandatory in this architecture. From there, the itinerary itself anticipates an evolution in two phases: first the sequential CFET technology will arrive, and later the united CFET structures, already in the A3 generation of 2038. The most interesting thing about this approach is that it changes the meaning of the Moore’s Law. IMEC recognizes that the contact poly pitch It will barely move between the A10 and A5 generations. Carry stuck at 42nm for several years. The density gains that we traditionally measured in individual transistor nanometers now depend on the height of the cell and how many layers can be stacked vertically. It is, in a way, a confession: horizontal scaling is exhausted, but the industry has found a third dimension in which it can continue growing. Image | Samsung More information | Tom’s Hardware In Xataka | US adds Netherlands to Pax Silica alliance: ASML host nation enters anti-Chinese bloc

produces for the first time the ultrapure silicon that its competitors controlled

China is a powerhouse in quantum technologies. In fact, in this area he looks directly at the US. These two countries have accumulated notable milestones over the past few years, including the quantum supremacybut his leadership diverges depending on the area. In quantum computing The US accumulates 34% of the most cited articles compared to 16% Chinesewhile in quantum telecommunications China leads with 34% of reference scientific production vs. 17% United States. The patent landscape reflects that same duality. China monopolizes around 60% of all requests global quantum technology patents registered in 2024, but when the filter is applied to international patents (those protected simultaneously in several countries), The US leads with 48% compared to 11% Chinese. Although, as we have just seen, China is making important achievements in the field of quantum technologies, it is to some extent dependent on some innovations from abroad. In the current situation of tension with the United States and its allies, this gigantic Asian country needs to become independent from foreign quantum technologies, and has just taken a very important step forward in this direction. The Chinese State Nuclear Corporation has announced that one of its research institutes has for the first time produced high-purity silicon-28 with an isotopic abundance greater than 99.99%. This isotope is decisive because it is considered the raw material for quantum hardware with the greatest scale potential: quantum computers based on silicon. Silence is everything The reason why silicon-28 matters so much must be found in physics. Natural silicon is composed mostly of silicon-28, but contains approximately 4.7% silicon-29, an isotope with non-zero nuclear spin. This detail is innocuous for a classical transistor, but is devastating for a quantum computer because the silicon-29 cores act as sources of magnetic noise that disturb the qubits, causing them to lose their quantum state in fractions of a second. Enriching the silicon until that impurity is almost completely eliminated is equivalent to building an anechoic room for the qubits: an environment where noise disappears and quantum information can persist long enough to operate on. A silicon-based quantum computer uses electron spin trapped in quantum dots (nanostructures manufactured on a silicon wafer) as a basic unit of information. Coherence, that is, the ability of the qubit to maintain its quantum state long enough to operate on it, depends directly on how quiet the surrounding environment is. In natural silicon the magnetic field generated by the silicon-29 nuclei drastically limits coherence In natural silicon, the magnetic field generated by the silicon-29 nuclei drastically limits this coherence. However, in silicon-28 enriched to 99.99% that field disappears. The most recent experiments on wafers manufactured with standard industrial CMOS processes have yielded coherence times of up to 803 microseconds and relaxation times of 6.3 seconds, as well as operating fidelities exceeding 99% in both one- and two-qubit gates. These numbers make sense when put against those of other qubit technologies. Superconducting qubits, the bet of Google and IBM, are extraordinarily fast in the execution of logic gates, but require cooling at temperatures close to absolute zero (around 15 millikelvin) and their coherence times They are measured in tens of microseconds. Trapped ion qubits, for their part, achieve coherences of several seconds and very low error rates, but their operations are slow and scaling them beyond a hundred qubits is extraordinarily difficult. In fact, they require ultra-fine vacuum systems, complex electromagnetic traps and surgically precise lasers. Spin qubits on silicon-28 are not the fastest or most coherent in absolute terms, but they combine sufficiently long coherence with an advantage that no other platform can match. That advantage is compatibility with the semiconductor industry as it exists today. Spin qubits in silicon are manufactured on 300 mm wafers using standard CMOS processes, the same ones that TSMC, Intel or Samsung use to produce classic transistors. This means that, in principle, the transition from a handful of qubits to millions integrated on a single chip is not limited by unknown physics, but by the same manufacturing engineering that the industry has mastered for decades. For China, which has invested many resources in building and strengthening its own semiconductor industry, being able to autonomously produce the base material on which this architecture depends is not only a scientific milestone; It’s a strategic move that connects its quantum agenda with its chip industry. And, as a tip, also with its broader objective of technological autonomy compared to the West. Image | Maxence Pira (Unsplash) More information | SCMP In Xataka | 38% of AI experts in the US have been trained in China. They are essential to sustain your leadership

As Silicon Valley perpetuates its workday, the four-day work week has found an unexpected ally: OpenAI

While in the mecca of the technology industry celebrates the “996” model (from nine in the morning to nine at night, six days a week) as a mantra to not to be left behind In the AI ​​race, the creator of ChatGPT stands out by proposing just the opposite: reducing working hours with a four day work week. OpenAI just published your report ‘Industrial Policy for the Intelligence Age: Ideas to Keep People First‘. In it, the company suggests that AI can be the perfect excuse for us to work fewer hours a week without losing a cent of our salary. The idea is not just an academic conjecture, but proposes a package of labor policies designed for the age of AI. Four-day days without touching the salary. One of the most surprising sections of the report refers to “efficiency dividends.” With them, OpenAI proposes that governments, companies and unions promote pilot tests of 32-hour days or four days of work per week without salary reduction, as has been established tested successfully in different countries around the world. The stated objective is to maintain the same levels of production and service, taking advantage of the automation options provided by AI and then making the leap to a model of permanent reduced working hours or cumulative vacation days for employees. The striking thing about the proposal is not its content itself, something that has already been implemented with success in some companiesthe key is who proposes the change. Instead of a union or a workplace welfare study, the idea comes from the company itself that is accelerating the transformation of the labor market around the world. Not just reduction in working hours: better pensions and care. OpenAI presents this measure as a way to redistribute part of the productivity benefits extra generated by AI, so that profits are not concentrated only in the shareholders or in the big technology companies, but that the entire population participates in this advance. The four-day week is just one of the most striking measures, but the report goes much further. OpenAI suggests that companies that profit from AI also increase their contributions to their employees’ pension plans (not just those of their managers as a bonus), and that they cover more of their employees’ healthcare expenses. He also proposes what he calls “benefit bonuses“, direct bonuses linked to improved productivity and subsidies for the care of minors and the elderly. If robots work, let them quote. The document recognizes that AI automation can lead to the massive displacement of jobs and further concentrate wealth in a very small number of large companies. That is why it calls for more robust social protection networks. Curiously, OpenAI’s postulate coincides with the statements made a few weeks ago did Bill Gatesarguing that if AI was to reduce dependence on human labor, taxation should shift from wages and contributions to capital gains and corporate profits. The document introduces the idea of ​​”taxes on automated work”, linked to jobs previously done by people who would be replaced by robots. In Xataka | The war in Iran has achieved something that no government has achieved: giving reasons to bring back teleworking Image | Unsplash (Nathan Kuczmarski)

The most opaque business in Silicon Valley has just published its best results. This is exactly what Palantir sells

Palantir has published some quarterly results that have surprised even its most optimistic analysts: revenues of $1.63 billion in the first quarter, 85% more than a year before. The company has also raised its annual forecast to almost 7,660 million. These are numbers that place Palantir in another league. And yet, many people don’t know exactly what they do. That is not an accident when it comes to this company, with such a specific type of activity. The context. Palantir has been building data analysis software for more than twenty years for governments and institutions that prefer not to make the headlines: the CIA, the FBI, the Pentagon, the United States immigration services… The company was founded in 2003 by Peter Thiel, the investor who also put money into Facebook and which today orbits in the power constellation. Trumpian, and by Alex Karp, its CEO, a notably eccentric figure with a PhD in Social Theory from Frankfurt. The combination of American intelligence money (In-Q-Tel, the CIA’s venture capital fund, was one of its first investors) and German university campus philosophy perfectly defines the moral ambiguity in which the company lives. In detailand. Palantir’s business has two legs. The first, and the one that is growing the most, is the American government: 687 million dollars in this quarter alone, 84% more than the previous year. The second leg is the commercial business with private companies, which has grown even faster (133%) to 595 million. But to understand how Palantir makes money you have to understand what it sells: Palantir Gotham It is its star product for governments and defense. It integrates dispersed data sources (satellite images, interceptions, movement logs, social networks, intelligence databases, etc.) and turns them into a coherent map that an analyst can interrogate. That is, it transforms oceans of noise into manageable information environments on which to make decisions. The screenshot that heads this article is an example. Palantir Foundry It is the business version. It does the same thing but for large companies: it unites data from different departments, cleans the information and allows automated workflows to be built on it. Maven AI It is their most recent and most controversial product. It is a command and control system that analyzes battlefield data and identifies targets in real time. The Pentagon is in the process of making it an official program of the American army, which would guarantee succulent long-term contracts. Between the lines. CEO Alex Karp This week he addressed his shareholders to explain to them that “the United States remains the constant core of the business. And that business is exploding.” Palantir’s rise is directly linked to increased defense spending, escalating geopolitical conflicts, and the growing use of AI in military contexts. In other words: when the world becomes more dangerous, Palantir makes more. Its business model is, to some extent, a barometer of global tension. Yes, but. Palantir Stock fell 1.5% in the aftermarket despite the good results, and they have accumulated a drop of around 18% so far this year. Investors have two questions without clear answers. The first: is 85% growth sustainable? The second, the most uncomfortable: what happens if the administration changes, if defense priorities change or if Congress tightens spending? A company whose main engine is a single client (the US government) has a concentration of risk that does not appear in the metrics they boast about. The money trail. The perennial debate about Palantir is not the financial one but the ethical one. The company has been at the center of some controversies for its work with ICE (the American immigration service) in identifying undocumented people, and for the role that its tools have played in military operations in different parts of the world. Karp does not shy away from these questions: he openly argues that the West needs companies willing to do this work, and that those who refuse simply leave the field open to others. It is an argument that its investors accept without many questions. And the results, for now, prove them right. In Xataka | AI is crucial for the US military. So he’s naming OpenAI and Palantir leaders as lieutenant generals Featured image | Palantir, Xataka with Mockuuups Studio

The Musk-Altman trial, the soap opera of the year in Silicon Valley

That Elon Musk and Sam Altman hate each other It’s nothing new. The news is that their rivalry is about to reach the courts. Yesterday jury selection took place which will decide whether, as Musk says, OpenAI violated its founding agreement to be a nonprofit organization. Much more is judged than that; The case aims to be a soap opera in which years of personal conflict will be aired. The accusation. In 2024, Elon Musk sued Sam Altman and OpenAI for having broken the company’s founding agreement, which said that “OpenAI is a non-profit artificial intelligence research company.” The CEO of SpaceX, who was one of the founders of OpenAI, claims that he was scammed and as soon as they got his money ($38 million) “he radically changed the narrative and dedicated himself to profiting (…) taking advantage of his humanitarian concern.” Musk calls for the dismissal of Sam Altman and his president, Greg Brockman, as well as $134 billion that would go to the nonprofit arm of OpenAI. OpenAI’s response. According to the company, Elon Musk was aware of the plan become a for-profit entitywhich was a necessary step for the company. OpenAI says that Musk is jealous and “regrets having retired” in 2018. In addition, they say that the money Musk contributed was not an investment, but a donation, and does not give him any ownership rights over OpenAI, as they say in Guardian. The breakup. OpenAI was founded in 2015 and shortly after, in 2017, it became clear that developing the dreamed-of AGI was going to require a lot of money, something difficult to achieve as a non-profit company. Here the option of a collaboration between OpenAI and Tesla was raised to solve the financing problem. According to OpenAI, Musk wanted full control of the company and this is where everything went wrong. The timing. Although he left the company in 2018, Musk does not sue OpenAI until 2024. During this time, OpenAI launched ChatGPT and ended up signing its great agreement with Microsoft. Apparently, this was what finally pissed off Musk, who accused OpenAI of having become a branch of Microsoft. Their objective was to annul the agreement, although much has happened since then. Today, OpenAI is the most valuable private company in the worldwith investments from giants such as NVIDIA, Amazon or Softbank. If Musk wins, his problem is no longer canceling the agreement with Microsoft, but something much more important: it puts at stake its planned IPO. A personal rivalry. Beyond the official accusations, Musk and Altman have a personal rivalry that has been brewing for years with countless public taunts and accusations. Musk has said that Altman is a liar and a scammer (He usually calls it ‘Scam Altman’) or that ChatGPT is a ‘woke’ AI. Altman has not been left behind; He has said that Musk is an idiot and even ridiculed him for the Tesla he bought and that was never deliveredgoing so far as to publish emails asking for a refund. In the trial, private details such as Elon Musk’s use of ketamine are expected to be aired, something that according to OpenAI is relevant since it portrays his mental state quite well. Let the show begin. Cover image | Village Global and Gage Skidmorevia Flickr In Xataka | MACROHARD is Elon Musk’s mockery of Microsoft: an AI agent that no other company can get. Still

Within Meta there is a race to see which employee consumes the most AI tokens. It’s the ‘Tokenmaxxing’ of Silicon Valley

There is a battle within Meta: see who spends the most AI tokens. This is the basic unit that AI uses to understand the language with which we order actions. It is like the “bridge” between our words and the numbers that the machine can process and, therefore, when ChatGPT either Google They present a model, they brag about the millions of tokens they can process. But tokens are also becoming a ‘spending’ unit in AI companies. Silicon Valleyso much so that they may be generating a toxic work culture. And Meta is an example of a company where employees compete to see how many tokens they can consume to become a Token Legend. Tokenmaxxing. It is not the first time that we talked about this. A few days ago, Jensen Huang -CEO of NVIDIA and one of the main instigators of this phenomenon- commented that he would be worried if an engineer who earns $500,000 did not spend at least $250,000 a year on tokens. Because tokens cost money and NVIDIA is already considering offering tokens as part of the signing bonuses for its artificial intelligence engineers. Goals. As it could not be otherwise, Meta does not want to miss this party. The company, which changed its name when the metaverse was going to be the big thing and, after the swerveis defined as a “native AI company”, is one of those that promotes its artificial intelligence engineers to keep a count of the tokens spent during their day. There is no official data, but there are reports revealed to media such as Business Insider and The Information which point out that some of these teams have very specific objectives related to the use of tokens. For example, the company expects 65% of its engineers to write more than 75% of code using AI tools by the middle of this year. The Scalable Machine Learning division has another objective, and so on in each of the code-related departments within Meta. Legend Token. In The Information, they directly point out that there is an internal classification table created by the employees themselves to gamify the work. It shows the 250 most intensive AI users in their tasks with an easy premise: the more tokens you spend, the more you climb in the ranking. The winner of this particular competition takes the title of ‘Token Legend’, or ‘Legend of Tokens’. It is turning an expectation into a kind of internal sport. The first paragraph of this article converted to tokens crazy spending. If we put the first paragraph of 542 words in the tool ‘tokenizer‘ from OpenAI, we see that that simple phrase has already consumed 121 tokens. Well: according to The Information, in the last 30 days the total token panel usage of that internal table was more than 60 billion (of ours) tokens And even if they want to dress it for sports and competition, it is still obligatory. In late 2025, Meta launched the ‘Level Up’ program where employees who complete the most tasks using AI earn badges. And more important than this: it made the use of AI a central criterion in its employee performance evaluations. This, obviously, sets salary and promotion objectives. Doubts. But of course, beyond paying to work, there are other underlying issues. One of the criticisms of this tokenmaxxing system is that AI companies like Meta or NVIDIA encourage spending more on tokens because, in this way, their own employees become consumers of the product they are creating. An easy example that software engineering analyst Gergely Orosz exposed which is as if Tim Cook, CEO of Apple, said that if one of his employees who earns $500,000 a year did not spend $50,000 on purchases in the App Store, he would be worried. Orosz continuous stating that productivity should not be measured in tokens spent, but in the results obtained. Industry issue. In any case, Meta and NVIDIA are not the only ones that measure their employees by their consumption of AI at work. It is something that is soaking in other AI majors, turning the tokens into an extra work benefit incorporated into the engineers’ remuneration wheel along with the base salary, performance bonuses and shares. HE esteem that an OpenAI engineer can process 210 billion tokens in a week and there are Claude Code engineers who accumulate more than $150,000 in tokens in one month. Basically it is merging part of your salary into the company that pays you. And… have they said anything from Meta? Yes, it’s not about volume, but about quality, pointing that performance rewards are based on the impact of the work and not the raw use of AI. Image | ‘Wolf of Wall Street’, Meta Logo. Edited In Xataka | Google Earth shows the world. The Spanish Xoople wants AI to understand it

Everyone in Silicon Valley has sided with the Trump administration. Everyone except Google DeepMind’s chief scientist

The big technology companies have changed their political color and the leaders who once boasted diversity and inclusiontoday they position themselves alongside Trump and they even have dinner with him. In Silicon Valley, almost no one talks about politics, much less to criticize the president, no one except one man. Jeff Dean. He is the chief scientist of Google DeepMind and one of the few voices in Silicon Valley that has a critical stance towards the Trump government and most importantly: he says it in public. As they say in the Wall Street JournalDean often shares messages critical of the current administration in your X accountwhere he has almost 430,000 followers, and has strongly condemned acts like the murder of Alex Pretti at the hands of ICe agents in the Minnesota protests. Support Anthropic. Jeff Dean was one of 30 Google and OpenAI employees who signed the letter of support towards Anthropicafter it sued the Pentagon for declaring them “a risk to national security”. The judge who agreed with Anthropic and blocked the decisioncited this letter of support as part of the context of the case. Other companies like Microsoft They also took a stand in favor of Anthropic, but they did so through a spokesperson. The rarity. Positions like that of Jeff Dean, who openly criticizes the government and also does so in a personal capacity, would not have stood out a few years ago. In 2018 technology companies turned to the #MeToo movement and also supported the mobilizations of the ‘Black Lives Matter’ movement. This happened during Trump’s first term, but during his second, figures like Jeff Dean have become a rarity in Silicon Valley. The second mandate. Mark Zuckerberg, Jeff Bezos, Sam Altman, Sundar Pichai, Satya Nadella, Tim Cook… the support of the technological leaders was noted from the investiture ceremony itself, which was not only attended but They donated a million dollars each and, as a reward, saved billions in taxes. Companies began to take measures immediately, such as remove from Meta’s anti-fake-news system either dismantle diversity, equity and inclusion departments. The delicate line. That Donald Trump is a fairly volatile leader is nothing new. Making him angry is as easy as contradicting him and his reaction can be excessive, And if not, tell Anthropic. In this sense, the fear of the possible consequences of a disagreement plays a role in this submission, because let us not forget that the government is the one who dictates export-import regulations and can make their business much easier. In the end, what is happening in Silicon Valley It is not so different from the situation in China that is so criticized from the American side. In Xataka | The war between Anthropic and the Pentagon points to something terrifying: a new “Oppenheimer Moment” Image | Wikipedia

China makes them almost as efficient as silicon ones

For years, polymer solar cells (popularly known as organic or “plastic” plates) have promised a real revolution in the renewable energy sector. Being light, flexible and even printable, their potential seemed limitless. However, in practice they had a big Achilles heel: they degraded quickly when exposed to air and their capacity to generate energy was far below that of the classic and heavy silicon panels. They were, in the eyes of the industry, almost a laboratory toy. But this narrative has just taken a historic turn. A team of scientists has managed to simultaneously overcome the barrier of performance and degradation, finally bringing these flexible plates closer to their long-awaited large-scale commercialization. A milestone that comes from China. Until today, manufacturing flexible solar panels meant taking a toll: either you lost efficiency or the material degraded quickly in the open air. Researchers at Wuhan University of Technology just broke that rule. Its new polymer cell reaches an efficiency of 19.1% – close on the heels of commercial silicon – and, above all, solves the problem of wear. As endorsed by the scientific journal matterthe device supports more than 2,000 hours of outdoor operation while retaining 97% of its initial capacity. In technical jargon, they have achieved a “T97 lifetime”, a metric that definitively takes this technology out of the experimental phase. The step definitive towards marketing. In statements to the magazine PV MagazineTao Wang, co-author of the research, highlights the magnitude of the finding: the stability demonstrated in these 2,000 hours allows us to extrapolate a useful life of the device that would exceed 100,000 hours of operation. Furthermore, this breakthrough puts an end to organic photovoltaics’ historic dilemma of the war between “efficiency vs. stability.” As the research indicatesUntil now, polymers (formed by long molecular chains) were very thermally stable and flexible, but inefficient; On the contrary, the “small molecules” were more efficient but too fragile and tended to crystallize over time, ruining the plate. This new development manages to combine the best of both worlds. The “invisible comb” at the microscopic level. Therein lies the secret of its success. Wei Li, another of the study’s lead authors, explains in PV Magazine that polymers have a mechanical problem: their long molecular chains tend to tangle, forming “disordered aggregates.” That disorder not only blocks the flow of electricity (reducing efficiency), but it exposes weak chemical bonds that accelerate the degradation of the board in sunlight. To solve this, the Wuhan team applied a strategy that was as elegant as it was effective: they introduced a small fraction of “small acceptor molecules” (SMA) into the polymer matrix. According to the studythis mixture acts as an invisible comb that “untangles” the long chains of the polymer, forcing them to pack together in a linear and orderly manner. This reduces empty spaces in the material, creating direct “highways” for electricity to flow without being lost, boosting efficiency and stopping photochemical deterioration in its tracks. A high-tech “sandwich”. For this chemical cocktail to work, the design of the plate was not left to chance. The cell was literally built like a sandwich on a microscopic scale. Instead of complicated heavy metal alloys, they used a transparent base on which they applied several ultrathin layers: one that captures light (the improved polymer), others that act as guides so that electrons do not escape and, finally, a very thin layer of silver to conduct electricity. The whole set results in a high precision, but extremely light device. And what does all this mean for the average user? According to the portal Interesting Engineeringthese findings pave the way to integrate highly efficient panels into tents, backpacks, clothing or covering the curved facades of buildings, without having to support the immense weight of silicon. This vision of the future is already taking its first commercial steps. As we saw a year ago at CESbrands like Anker Solix are already experimenting with prototypes of jackets that integrate solar panels and power banks to keep a mobile phone charged, or beach umbrellas capable of charging a portable refrigerator using continuous photovoltaic cells. The difference is that, thanks to the new molecular advances achieved in China, this “wearable” and portable self-consumption technology will take a brutal leap: it will be much more stable, durable and easier to mass produce. The future is already flexible. The absolute hegemony of silicon – rigid, heavy and with a high manufacturing energy cost – is beginning to have a real alternative on the horizon. Research from Wuhan University of Technology shows that understanding and manipulating how molecules behave and intertwine was the master key to getting organic technology out of the laboratory. The future of solar energy no longer only seeks to be efficient; Now it is ready to be flexible, ultralight and, finally, durable. Image | RawPixel Xataka | Solar panels have an invisible and very brief moment in which they do not work. And solving it is key to your future

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.