According to the source, an X post dated Sep 22, 2025 claims MIT's SCIGEN and Samsung's PaRS embed physics into AI so LLMs can propose exotic yet feasible materials for quantum tech, energy, and chips; source: public X post dated Sep 22, 2025. As of the assistant's knowledge cutoff, independent peer-reviewed papers or official releases describing 'SCIGEN' or 'PaRS' were not identified, so traders should wait for primary confirmation from MIT or Samsung before repricing related equities; source: MIT Newsroom and Samsung SAIT Newsroom checked through Oct 2024. Proven precedent supports the thesis that physics-informed ML accelerates materials discovery—DeepMind's GNoME predicted about 2.2 million crystal candidates with over 380,000 likely stable, compressing discovery timelines; source: Nature (2023) and DeepMind research blog (2023). If similar tools are validated, near-term watchlists include materials informatics vendors, semiconductor design and equipment leaders, and BTC mining hardware supply chains given potential efficiency gains from new thermal and substrate materials; source: Nature (2023) on GNoME implications and Cambridge Bitcoin Electricity Consumption Index on miner power economics (2023–2024).