[{"data":1,"prerenderedAt":30},["ShallowReactive",2],{"nr-en-alibaba-qwen-vier-supraleiter-ki-entdeckung":3},{"slug":4,"title":5,"dek":6,"date":7,"time":8,"publishedAt":9,"updated":10,"updatedAt":10,"dateFmt":11,"updatedFmt":10,"kind":12,"tier":13,"author":14,"authorName":15,"topics":16,"tracker":22,"trackerLabel":23,"headlineStat":24,"image":25,"ogImage":26,"imageAlt":5,"csv":10,"minutes":27,"words":28,"html":29},"alibaba-qwen-vier-supraleiter-ki-entdeckung","Alibaba discovers four new superconductor candidates with AI agent – verified in lab","Alibaba's DAMO Academy used the AI system Elements Clo to identify and experimentally confirm four previously unknown superconductor candidates. Evidence that AI in materials research is moving far beyond text generation.","2026-07-06","09:26","2026-07-06T09:26:00+02:00","","July 6, 2026","news","standard","ideal-syka","Ideal Syka",[17,18,19,20,21],"AI research","materials science","superconductors","Alibaba","fundamental research","\u002Fstand-der-ki","AI progress","4 new superconductor candidates verified","\u002Fnewsroom\u002Fimg\u002Falibaba-qwen-vier-supraleiter-ki-entdeckung.webp","\u002Fog-nr\u002Falibaba-qwen-vier-supraleiter-ki-entdeckung.en.png",2,470,"\u003Cp>Alibaba's research division DAMO Academy has developed an AI agent called \u003Cstrong>Elements Clo\u003C\u002Fstrong> that discovered four new superconductor candidates and verified them in experiments. This demonstrates that artificial intelligence is becoming a genuine tool in fundamental research – not just for code and chat.\u003C\u002Fp>\n\u003Ch2>The essentials\u003C\u002Fh2>\n\u003Cul>\n\u003Cli>\u003Cstrong>Elements Clo\u003C\u002Fstrong> is described by DAMO Academy as the industry's first AI agent for superconductor discovery, based on a \u003Cstrong>1-billion-parameter model\u003C\u002Fstrong> trained on \u003Cstrong>125 million\u003C\u002Fstrong> molecular and crystal structures\u003C\u002Fli>\n\u003Cli>The agent screened \u003Cstrong>2.4 million\u003C\u002Fstrong> stable crystal structures in \u003Cstrong>28 hours\u003C\u002Fstrong> of GPU computing and narrowed them to approximately \u003Cstrong>68,000\u003C\u002Fstrong> promising candidates\u003C\u002Fli>\n\u003Cli>\u003Cstrong>Four previously unreported compounds\u003C\u002Fstrong> were experimentally confirmed after further analysis\u003C\u002Fli>\n\u003Cli>Renmin University of China and the University of Chinese Academy of Sciences participated in the research\u003C\u002Fli>\n\u003C\u002Ful>\n\u003Ch2>How AI accelerates the research process\u003C\u002Fh2>\n\u003Cp>Superconductors are materials whose electrical resistance vanishes in extremely cold environments and which expel magnetic fields. These properties are considered key for future technologies such as more efficient power grids, quantum computing, and magnetic levitation trains. Until now, discovering new materials has been a time-consuming trial-and-error process – because the theory of superconductivity remains incomplete.\u003C\u002Fp>\n\u003Cp>Elements Clo changes this dynamic: the system automatically analyzes scientific literature and crystal data to identify materials with high probability of superconductivity. Instead of researchers manually sifting through millions of candidates, AI handles the pre-filtering. Scientists then focus on experimentally validating the most promising candidates.\u003C\u002Fp>\n\u003Ch2>A new paradigm for AI in science\u003C\u002Fh2>\n\u003Cp>The project illustrates a broader trend: tech companies are no longer limiting AI to language generation or software development. They are expanding its use to literature review, data analysis, and hypothesis formulation for experimenters.\u003C\u002Fp>\n\u003Cp>In materials science especially, AI could have enormous impact. The search for new materials requires reviewing countless compounds – a classic bottleneck. The established superconductor database \u003Cstrong>SuperCon\u003C\u002Fstrong> contains only about \u003Cstrong>2,000\u003C\u002Fstrong> registered materials. Elements Clo demonstrates how AI can dramatically shorten this search phase.\u003C\u002Fp>\n\u003Cblockquote>\n\u003Cp>"Elements Clo handles paper searches, structure screening and candidate compression, while researchers focus on experimental verification,"\u003C\u002Fp>\n\u003C\u002Fblockquote>\n\u003Cp>DDAO Academy describes the division of labor. This model could transfer to other research fields – anywhere large datasets and theoretical uncertainty converge.\u003C\u002Fp>\n\u003Ch2>What this means for German research\u003C\u002Fh2>\n\u003Cp>For German companies and research institutions, there is an important message here: AI is becoming a tool for fundamental research and materials development. Those who do not build or integrate such systems risk falling behind in critical technology fields. At the same time, an opportunity emerges: German expertise in materials science, chemistry, and physics could be multiplied with AI tools. The question is whether domestic institutes and corporations actively pursue this path – or leave the field to Alibaba, OpenAI, and others.\u003C\u002Fp>\n\u003Ch2>Sources\u003C\u002Fh2>\n\u003Cul>\n\u003Cli>\u003Ca href=\"https:\u002F\u002Fwww.digitaltoday.co.kr\u002Fen\u002Fview\u002F78339\u002Falibaba-ai-finds-four-previously-unknown-superconductor-candidates-verified-in-experiments\">디지털투데이\u003C\u002Fa>\u003C\u002Fli>\n\u003C\u002Ful>\n\u003Cp>\u003Cem>Editorially owned by \u003Ca href=\"\u002Fen\u002Fautor\u002Fideal-syka\">Ideal Syka\u003C\u002Fa>. Sources and method: \u003Ca href=\"\u002Fen\u002Fredaktion\">Newsroom & method\u003C\u002Fa>. Tips and corrections: \u003Ca href=\"mailto:ai@i6eal.de\">ai@i6eal.de\u003C\u002Fa>.\u003C\u002Fem>\u003C\u002Fp>\n",1783332802352]