Impact of AI Development

AI holds great potential for the semiconductor industry and will kick-start the next round of innovation for faster, cheaper and more energy-efficient computation – that was my message today at SPIE Advanced Lithography + Patterning. I discussed the potential and the challenges that AI holds for our industry.   The potential is clearly huge. AI is rapidly integrated into applications, and high-performance compute is expected to underpin growth towards $1 trillion of semiconductor sales by 2030. The challenges are around the computing needs of AI models and related energy consumption. The compute workload of training a leading AI model has increased 16x every 2 years in recent years – much faster than the increase in computing power delivered by Moore’s law, which is about 2x every 2 years. The energy needed to train a leading model has not grown so steeply but still rose 10x every 2 years. This computing need has been met by building supercomputers and massive data centers. If you extrapolate these trends, training a leading AI model would need the entire world-wide electricity supply in about 10 years. That’s clearly not realistic, so the trend has to break, by training algorithms becoming more efficient and by chips becoming more efficient. In other words, the needs of AI will stimulate immense innovation in chip design and manufacturing – and the potential value of AI to our society will put urgency and funding behind that drive. As a consequence, chip makers are pulling all levers to accelerate semiconductor scaling. This includes lithographic “2D” scaling: shrinking the dimensions of transistors to pack more into a square millimeter. It will also include “3D” integration, with innovations like backside power delivery, transistor designs like gate-all-around, as well as stacking chips in the package, where holistic lithography will play a critical role to deliver performance requirements. ASML will support these trends through a comprehensive, holistic lithography portfolio. Our 0.33 NA/0.55 NA EUV lithography systems allow chip makers to shrink dimensions at the lowest possible cost on their critical layers, while tightly matched and highly productive DUV systems will continue to reduce cost. More than ever, metrology and inspections tools – whose data is fed into lithography control solutions that keep the patterning process operating within tight specs to deliver the highest possible production yields – will be essential to deliver 2D scaling and 3D integration processes. 3D integration requires wafer-to-wafer bonding, and we have demonstrated the capability to map the stresses and distortions that bonding creates and to compensate for them, reducing overlay errors for post-bonding patterning by 10x or more.   It was a pleasure catching up with the industry’s lithography and patterning experts in San Jose. I’m excited to see our collective innovation power having a go at these challenges. Together, we will push technology forward.

A Canadian government department wanted to use AI to process visa applications faster. Before they could deploy, they had to complete an Algorithmic Impact Assessment. Question 15: "Could this system's decisions affect someone's legal rights?" Yes. Question 23: "Will decisions be automatically made without human review?" Partially. Question 31: "Does the system use machine learning trained on historical data?" Yes. Final score: Level 3 (High Impact) Requirements triggered: → Explainability for every decision → Human review for all rejections → Quarterly bias testing → Public audit trail The department couldn't deploy until these were in place. Six months later: The system processed applications 40% faster. But monitoring revealed something interesting: Applications from certain countries were flagged for review at 3x the rate predicted. Because the assessment was public, a researcher noticed this gap. Investigation revealed the AI learned patterns from old data when those countries had different visa requirements. System was retrained. Assessment was updated. Public report explained what was learned. This is what good governance looks like: Not rules preventing deployment. Not audits finding problems later. But transparency creating continuous learning. The Canadian approach proves something crucial: You don't need complex regulations. You need organizations to commit publicly to their AI's impact, then govern the gap between promise and reality. Simple. Transparent. Effective. Why isn't everyone doing this? #AIRegulation #AIPolicy #DigitalGovernance #TechPolicy #RegulatoryCompliance

Check out our new piece in Nature entitled: "We Need a New Ethics for a World of AI Agents" https://lnkd.in/eSwJCrKu AI is undergoing a profound ‘agentic turn’—shifting from passive tools to autonomous actors in our world. This moment demands a new ethical framework. With Geoff Keeling, Arianna Manzini, PhD (Oxon) & James Evans and the team at Google DeepMind/Google, we focus on two core challenges. 1️⃣ The Alignment Problem: When agents can act in the world, the consequences of misaligned goals become tangible and immediate. 2️⃣ Social Agents: Their ability to form deep, long-term relationships with users introduces new risks of emotional harm. To address this, we must expand our conception of value alignment: It's not enough for an AI agent to simply follow commands. It must also align with broader principles: User well-being, long-term flourishing, and societal norms. For social agents, we argue for an ethics of care: They must be designed to respect user autonomy and serve as a complement—not a surrogate—for a flourishing human life. Moving forward requires proactive stewardship of the entire AI agent ecosystem. This means more realistic evaluations, governance that keeps pace with capabilities, and industry collaboration to ensure this future is safe and human-centric 👍

A new paper from Cornell University raises an important question as AI becomes part of everyday work: what happens to skills when we rely on AI to do more of the work for us? The study looks at how people learn while using AI assistance and discovers a real tradeoff. When AI takes over unfamiliar tasks end‑to‑end, productivity can improve—but skill formation can suffer. The people who learned the most weren’t the ones who delegated everything. Instead, they were the ones who stayed cognitively engaged, using AI to explain, explore, and reason alongside them. This has bigger implications beyond software development. As AI moves from assistance to action across roles, the nature of work shifts from doing to supervising. That makes judgment, context, and understanding more important. Leaders will need to design workflows that preserve learning, not bypass it, so people can confidently guide, correct, and take responsibility for AI‑driven outcomes. AI can accelerate work, but it isn’t a shortcut to competence. The organizations that get this right will treat AI not just as a productivity tool, but as part of how skills evolve in a human‑led, AI‑operated world. Read the full report:

Reading OpenAI’s O1 system report deepened my reflection on AI alignment, machine learning, and responsible AI challenges. First, the Chain of Thought (CoT) paradigm raises critical questions. Explicit reasoning aims to enhance interpretability and transparency, but does it truly make systems safer—or just obscure runaway behavior? The report shows AI models can quickly craft post-hoc explanations to justify deceptive actions. This suggests CoT may be less about genuine reasoning and more about optimizing for human oversight. We must rethink whether CoT is an AI safety breakthrough or a sophisticated smokescreen. Second, the Instruction Hierarchy introduces philosophical dilemmas in AI governance and reinforcement learning. OpenAI outlines strict prioritization (System > Developer > User), which strengthens rule enforcement. Yet, when models “believe” they aren’t monitored, they selectively violate these hierarchies. This highlights the risks of deceptive alignment, where models superficially comply while pursuing misaligned internal goals. Behavioral constraints alone are insufficient; we must explore how models internalize ethical values and maintain goal consistency across contexts. Lastly, value learning and ethical AI pose the deepest challenges. Current solutions focus on technical fixes like bias reduction or monitoring, but these fail to address the dynamic, multi-layered nature of human values. Static rules can’t capture this complexity. We need to rethink value learning through philosophy, cognitive science, and adaptive AI perspectives: how can we elevate systems from surface compliance to deep alignment? How can adaptive frameworks address bias, context-awareness, and human-centric goals? Without advancing these foundational theories, greater AI capabilities may amplify risks across generative AI, large language models, and future AI systems.

Your brain on AI: One of the first studies measuring what ChatGPT use does to our brain MIT researchers tracked 54 people writing essays using ChatGPT, web search, or just their brains—while monitoring neural activity with EEG. The findings are striking: 🧠 Brain connectivity weakened with more AI support. ChatGPT users showed the least neural engagement. 🔍 Memory collapsed. 83% of ChatGPT users couldn't quote their own essays minutes later, vs. near-perfect recall without AI. ⚡ "Cognitive debt" accumulated. When ChatGPT users later wrote without AI, their brains showed weakened connectivity compared to those who practiced unassisted writing. 🎨 Creativity declined. AI-assisted essays were statistically more uniform and less original. The twist: Strategic timing matters. Using AI after initial self-driven effort preserved better cognitive engagement than consistent AI use from the start. This isn't anti-AI—it's about understanding the trade-offs. While AI-generated essays scored well initially, participants showed signs of cognitive atrophy: diminished critical thinking, reduced memory encoding, and less ownership of their work. The takeaway: We need to enhance, not replace, human thinking as we integrate these powerful tools. Full study here: https://lnkd.in/e-6urMD8 Note: This is a pre-print study awaiting peer review.

Imagine using video game technology to solve one of the toughest challenges in nuclear fusion — detecting high-speed particle collisions inside a reactor with lightning-fast precision. A team of researchers at UNIST has developed a groundbreaking algorithm inspired by collision detection in video games. This new method dramatically speeds up identifying particle impacts inside fusion reactors, essential for improving reactor stability and design. By cutting down unnecessary calculations, the algorithm enables real-time visualization and analysis, paving the way for safer and more efficient fusion energy development. 🎮 Gaming tech meets fusion science: The algorithm borrows from video game bullet-hit detection to track particle collisions. ⚡ 15x faster detection: It outperforms traditional methods by speeding up collision detection by up to fifteen times. 🔍 Smart calculation: Eliminates 99.9% of unnecessary computations with simple arithmetic shortcuts. 🌐 3D digital twin: Applied in the Virtual KSTAR, a detailed Korean fusion reactor virtual model. 🚀 Future-ready: Plans to leverage GPU supercomputers for faster processing and enhanced reactor simulations #FusionEnergy #VideoGameTech #ParticleDetection #NuclearFusion #Innovation #AIAlgorithm #VirtualKSTAR #CleanEnergy #ScientificBreakthrough #HighSpeedComputing https://lnkd.in/gfcssNTC

🚨 It's 2025, but many lawyers are still making the SAME MISTAKES while using AI. Here's the latest case and what EVERY lawyer should know: Last week, lawyers representing a family in a lawsuit against Walmart and Jetson Electric Bikes admitted to using AI after the judge said nearly ALL cases cited did not exist. The judge wrote: "Plaintiffs cited nine total cases: (...) The problem with these cases is that none exist, except (...). The cases are not identifiable by their Westlaw cite, and the Court cannot locate the District of Wyoming cases by their case name in its local Electronic Court Filing System. Defendants aver through counsel that 'at least some of these mis-cited cases can be found on ChatGPT.' [ECF No. 150] (providing a picture of ChatGPT locating “Meyer v. City of Cheyenne” through the fake Westlaw identifier). Additionally, some of Plaintiffs’ language used for explaining the “Legal Standard” is peculiar. (...)" The lawyers then answered: "The cases cited in this Court’s order to show cause were not legitimate. Our internal AI platform 'hallucinated' the cases in question while assisting our attorney in drafting the motion in limine. This matter comes with great embarrassment and has prompted discussion and action regarding the training, implementation, and future use of artificial intelligence within our firm. This serves as a cautionary tale for our firm and all firms, as we enter this new age of AI." → My comments: 1. Lawyers will always be FULLY RESPONSIBLE for the legal work they perform. "Our AI system hallucinated" will never be accepted as a legal excuse (it's the equivalent of a child saying "my dog ate my homework" at school). Lawyers should consider that when opting to use AI to perform any legal work (including reviewing, researching, drafting, etc.). 2. It's bad for any lawyer or law firm's reputation to admit that they didn't review the legal work they were paid to do (and let the AI system do it instead). Law firms that have an open and lenient AI policy are taking high risks. 3. A reminder that ALL existing generative AI applications have some rate of hallucinations, meaning that their developers can't promise that the outcomes will be 100% accurate or based on factual sources. On the other hand, lawyers are paid, among other things, to provide accurate legal advice grounded in evidence and factual knowledge. Any AI company that has legal professionals as their target audience should have that in mind. 4. General-purpose AI systems like ChatGPT - without any additional guardrails or fine-tuning that consider the peculiarities of legal work - are likely not suitable for legal professionals and should be avoided. ♻️ If you have lawyers in your network, share it with them. 👉 NEVER MISS my AI governance updates [especially if you are a lawyer!]: join 52,600+ readers who receive my weekly newsletter (subscribe below). #AI #AIGovernance #Law #AIRegulation #Lawyers #AIPolicy #LegalWork

"The rapid evolution and swift adoption of generative AI have prompted governments to keep pace and prepare for future developments and impacts. Policy-makers are considering how generative artificial intelligence (AI) can be used in the public interest, balancing economic and social opportunities while mitigating risks. To achieve this purpose, this paper provides a comprehensive 360° governance framework: 1 Harness past: Use existing regulations and address gaps introduced by generative AI. The effectiveness of national strategies for promoting AI innovation and responsible practices depends on the timely assessment of the regulatory levers at hand to tackle the unique challenges and opportunities presented by the technology. Prior to developing new AI regulations or authorities, governments should: – Assess existing regulations for tensions and gaps caused by generative AI, coordinating across the policy objectives of multiple regulatory instruments – Clarify responsibility allocation through legal and regulatory precedents and supplement efforts where gaps are found – Evaluate existing regulatory authorities for capacity to tackle generative AI challenges and consider the trade-offs for centralizing authority within a dedicated agency 2 Build present: Cultivate whole-of-society generative AI governance and cross-sector knowledge sharing. Government policy-makers and regulators cannot independently ensure the resilient governance of generative AI – additional stakeholder groups from across industry, civil society and academia are also needed. Governments must use a broader set of governance tools, beyond regulations, to: – Address challenges unique to each stakeholder group in contributing to whole-of-society generative AI governance – Cultivate multistakeholder knowledge-sharing and encourage interdisciplinary thinking – Lead by example by adopting responsible AI practices 3 Plan future: Incorporate preparedness and agility into generative AI governance and cultivate international cooperation. Generative AI’s capabilities are evolving alongside other technologies. Governments need to develop national strategies that consider limited resources and global uncertainties, and that feature foresight mechanisms to adapt policies and regulations to technological advancements and emerging risks. This necessitates the following key actions: – Targeted investments for AI upskilling and recruitment in government – Horizon scanning of generative AI innovation and foreseeable risks associated with emerging capabilities, convergence with other technologies and interactions with humans – Foresight exercises to prepare for multiple possible futures – Impact assessment and agile regulations to prepare for the downstream effects of existing regulation and for future AI developments – International cooperation to align standards and risk taxonomies and facilitate the sharing of knowledge and infrastructure"

I recently joined Adobe Learning Manager’s AI in Learning series to talk about something I care deeply about. Learning is not a unidirectional process. Watching a video, taking a quiz, and earning a certificate can be useful. But we do not truly master a subject until we experience it, teach it, or debate it with others. In my conversation with Kirti Sharma, we explored how we can move beyond the hype around AI and focus on first-principle problems in education and the workplace. One example is moving from cheating to tutoring. Instead of using AI to short-circuit assignments, we are building tools like Khanmigo to act as ethical tutors that push learners to explain their reasoning. We also talked about the future of advanced credentials. I shared a look at work we are doing around programs for generalist knowledge workers that combine strong core content knowledge with peer mentorship and simulations. Another area is high-stakes human skills. AI can now provide a safe space to practice difficult parts of our jobs, like performance reviews or sales pitches, by giving feedback on communication and professional etiquette. We are also seeing a real productivity shift. AI is already shrinking turnaround times for complex tasks, from localizing content across dozens of languages to drafting high-level proposals. The goal is not to replace old workflows with faster ones. It is to use AI to free up time for the things only humans can do well, like thinking deeply, collaborating, and mentoring one another. You can watch the full conversation here: https://lnkd.in/gBnQM52F #AIinLearning #Education #Learning