- IoT For All
- Posts
- AI's taking out superbugs, one atom at a time
AI's taking out superbugs, one atom at a time
And our top AI story of the week
AI For All
August 15, 2025
Hello readers,
Welcome to the AI For All newsletter! Today, we’ll be exploring how AI has been instrumental in designing new treatments for antibiotic-resistant superbugs, the tipping point of AI-powered IoT implementation in logistics, and more!
AI In Action: Taking out superbugs, one atom at a time
Researchers at MIT have used generative AI to design two entirely new antibiotics capable of killing drug-resistant superbugs like MRSA and gonorrhea, as reported by the BBC. The AI was trained on tens of millions of compound structures, including ones not yet synthesized, and then designed molecules atom-by-atom, based on how different structures interact with bacterial targets. After lab and animal testing, two promising candidates emerged that killed the target bacteria, offering a proof-of-concept for a faster, cheaper path to novel antibiotics.
“AI can enable us to come up with molecules, cheaply and quickly and in this way, expand our arsenal, and really give us a leg up in the battle of our wits against the genes of superbugs.”
Generative models can explore vast chemical spaces far beyond what humans or traditional computational methods can cover. This dramatically shortens the early-stage drug discovery process, which is typically slow, expensive, and prone to failure. AI can design candidates from scratch to target specific infections, avoid toxicity, and even filter for manufacturability. This offers a powerful new tool to replenish the drug pipeline after decades of stagnation.
The newly-designed compounds will still require further refinement and years of clinical testing, but the success of this approach signals a major shift in the fight against drug-resistant infections. It shows AI is actively enabling breakthroughs that wouldn’t be feasible using conventional methods, potentially reviving antibiotic development by changing the economics of pharmaceutical R&D.
🔥 Rapid Fire
GPT-5 disappoints with troubled rollout and not much improvement
What if AI doesn’t get much better than this?
Companies report GenAI is still yet to pay off, ‘little effect’ to bottom line
GPT-5 is supposed to be ‘PhD level’ smart. It can’t even label a map.
The high costs and thin margins threatening AI coding startups
A flirty Meta AI bot invited a retiree to meet. He never made it home.
Why Elon Musk is threatening to sue Apple over Grok's App Store ranking
AI deepfakes in finance, 70% of new enrollments to some firms are fake
GPT-5 doesn't dislike you—it might just need an emotional benchmark
Apple reportedly planning AI comeback, complete with a tabletop robot
AI that perceives would experience time differently than humans
YouTube will try to estimate users’ ages with AI
📖 What We’re Reading
“After years of limited adoption due to implementation complexity and analytical constraints, the integration of AI processing capabilities — particularly through Model Context Protocol (MCP) servers — is fundamentally altering the technical and economic feasibility of comprehensive IoT deployments in supply chain operations.”