<h2>Must know</h2>
<ul>
<li><strong>Meta bets big on nuclear for AI power</strong>: Meta is securing up to 6.6 GW of nuclear energy capacity by 2035 through partnerships with Vistra, TerraPower, and Oklo. This massive investment aims to power its expanding AI infrastructure and data centers, particularly the Prometheus supercluster in Ohio. The move highlights the immense energy demands of advanced AI and Meta’s strategy to ensure a reliable, clean baseload for its compute needs, as reported by Meta itself. <a href="https://app.daily.dev/posts/elp6Y0W2r" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>India’s smartphone security demands spark industry outcry</strong>: India is proposing 83 new smartphone security standards, including source code access and 12-month device log storage, which major tech companies like Apple and Samsung are calling unprecedented and invasive. While the Indian IT ministry publicly denies seeking source code, Reuters reports contradict this with leaked government documents. This regulatory push aims to reduce reliance on foreign tech but faces strong industry pushback over commercial risks and technical feasibility. <a href="https://app.daily.dev/posts/C0BEamxdR" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Linus Torvalds embraces AI for coding</strong>: Even Linus Torvalds, the creator of Linux, is now using AI coding tools. He recently released AudioNoise, an open-source guitar pedal simulator, and used Google’s Antigravity AI-assist IDE (likely Gemini) for the Python visualization tool. Torvalds describes this ‘vibe coding’ as streamlining the process between searching for code examples and manual coding, marking a significant shift in AI adoption by a major open-source leader. <a href="https://app.daily.dev/posts/fe8LDMARY" target="_blank" rel="noopener nofollow">Read more</a></li>
</ul>
<h2>Good to know</h2>
<ul>
<li><strong>US and China in quantum and AI race</strong>: An NSF-funded ERVA report warns the US could lose quantum leadership to China and the EU without increased investment, while China’s AI sector is accelerating with successful IPOs and optimizing models for less capable hardware. <a href="https://app.daily.dev/feed-by-ids?id=2Pu5NXJ2N&amp;id=H3IqbB7DB&amp;id=I59qgstPX" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Instagram password reset chaos and data leak claims</strong>: Millions of Instagram users received unexpected password reset emails, and Malwarebytes claims to have found a 17.5 million account dataset on the dark web, though Meta denies any breach, attributing the emails to an external technical issue. <a href="https://app.daily.dev/posts/MEOecqk3z" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Universal Commerce Protocol and Amazon’s AI shopping stir controversy</strong>: Google, Shopify, and others launched UCP, an open-source standard for AI-driven commerce, while Amazon’s ‘Buy For Me’ feature is drawing criticism for automatically listing merchant products without consent, causing issues like incorrect orders and exposing wholesale prices. <a href="https://app.daily.dev/feed-by-ids?id=7CfzB00S6&amp;id=HH05G9maY&amp;id=MhDEwyYbE" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>D-Wave acquires Quantum Circuits for gate-model expansion</strong>: D-Wave Quantum is acquiring Quantum Circuits Inc. for $550 million to integrate annealing quantum computing with error-corrected gate-model technology, aiming for a commercially available gate-model system by 2026. <a href="https://app.daily.dev/posts/rdMlkyvEb" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>European Commission consults on open source strategy</strong>: The EU is consulting on an open source strategy to reduce reliance on US tech giants, aiming to classify open source as ‘critical infrastructure’ and foster EU-based alternatives in cloud, AI, and cybersecurity. <a href="https://app.daily.dev/posts/nSTd14JQu" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Nanolasers 3D printed directly on chips</strong>: KAIST and POSTECH researchers developed a direct 3D printing method for vertical nanolasers on semiconductor chips, bypassing traditional lithography and promising higher efficiency for optical computing and quantum communication. <a href="https://app.daily.dev/posts/a8cgcR8ai" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Redundant quantum information backup achieved</strong>: University of Waterloo researchers developed a method to back up quantum information redundantly by encrypting qubits into multiple entangled systems, allowing data storage across several servers without violating the no-cloning theorem. <a href="https://app.daily.dev/posts/cd0RSPfq1" target="_blank" rel="noopener nofollow">Read more</a></li>
</ul>
<h2>Bullshit police</h2>
<ul>
<li><strong>AI code generation: the new productivity paradox</strong>: Apparently, AI code generation tools are now creating 70-90% of production code, and if you’re not adapting, you’re ‘falling behind.’ The article, which conveniently lacks any actual data sources, suggests developers should challenge AI with ‘real’ tasks and create custom benchmarks instead of relying on vendor claims. Because nothing says ‘productivity’ like spending your time validating AI’s output and building benchmarks for tools that are supposed to make you faster. It’s almost like we’re just shifting the work, not eliminating it. <a href="https://app.daily.dev/posts/LeN1TdPNx" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>GenAI project failures: blame the infrastructure, not the hype</strong>: A seasoned practitioner claims 80% of GenAI projects fail, not because the models are overhyped, but due to a lack of ‘production-ready infrastructure.’ The solution? Seven ‘essential production folders’ like config management, prompt versioning, and error handlers. So, after all the talk of AI revolutionizing everything, it turns out the real problem is… we still need good software engineering practices? Who would’ve thought? It’s almost like throwing a fancy new model at a messy codebase doesn’t magically fix your architectural debt. <a href="https://app.daily.dev/feed-by-ids?id=ExsMXw27a&amp;id=yocwu10Ql" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>AI coding tools: the new focus fragmentation device</strong>: According to some unnamed ‘experts,’ those 5-30 second pauses while AI code generators do their thing are ‘destroying deep work’ and causing ‘context switching to distractions.’ Their brilliant solution? An ‘AI Detox’ protocol where you physically remove your hands from the keyboard and close distracting apps. Because the problem isn’t the AI, it’s your inability to stare blankly at a screen for half a minute without checking Slack. Maybe the AI should generate a focus app for us while it’s thinking. <a href="https://app.daily.dev/feed-by-ids?id=eC9HsHF69&amp;id=infazu71Y&amp;id=uTQFFroPQ" target="_blank" rel="noopener nofollow">Read more</a></li>
</ul>
