Laura N Montoya, CSM

We are far from one shot drug discovery in my opinion but we are palpably entering an age where AI is rapidly accelerating DBTL loops and meaningful solutions can be incepted in a handful of days 🔁 From a Bio x AI hackathon in Berlin this weekend organised by Nucleate, Project Europe and JUNI to one in SF in the next few weeks by Tamarind Bio, BioRender and Y Combinator 👾 It’s never been more exciting to be in this space! 🌱 #biohack #hackathon #aixbio

31

mpathic is becoming THE partner to the AI industry as they work to safeguard users from adverse mental health and safety outcomes As millions of users — including large numbers of young people — increasingly turn to AI chatbots as their first-line “counselors” and confidants, Seattle-based startup mpathic is stepping in to ensure those digital agents don’t provide dangerous advice when it matters most.

33

4 Comments

NYU Langone Health’s informatics leaders recently predicted fully autonomous clinical AI within 5 years with no physician sign-off. Algorithms titrating blood pressure meds, retinopathy screening running end-to-end without a clinician in sight. Their reasoning is hard to argue with as we're facing a shortage of 85,000+ physicians by 2036 and the current care model doesn't have an answer for that, and it seems clear that AI will be playing more of role in providing care going forward. But the next question is more difficult to answer: Are we automating the right things, for the right reasons?   Utah is already stress-testing this with Doctronic by automating routine prescription renewals for chronic condition patients. I wrote about this a month ago - renewals are repetitive, pulling clinicians away from patients who need actual judgment and create gaps in care when they fall through the cracks. If that pilot scales, it frees up real capacity. What makes it work is that prescription renewals are largely protocol-driven. The decision tree is narrow, the inputs are structured, and removing friction there doesn't mean removing judgment from care. Clay Christensen would have called these protocol-driven care pathways 'precision medicine,' having advanced originally from the primordial soup of 'intuitive medicine,' to the mid-stage 'empirical medicine,' to finally precision status. We're seeing the same logic play out across our portfolio companies. Janie automates the administrative workflows that quietly drain clinical capacity, things like billing, prior auth, and care coordination, giving practices back time that should never have been spent on paperwork. FlyteHealth takes a similar approach on the clinical side, combining a multidisciplinary care team with technology to manage obesity, hypertension, and diabetes at scale. The care team stays central and the automation handles everything that doesn't require them. That's the model I find most compelling. Where I get more cautious is with patients who have multiple conditions, polypharmacy risk, or subtle behavioral changes that don’t show up in structured data. A system that's right 95% of the time still needs airtight guardrails for the other 5%, and building those guardrails well requires people who understand what failure looks like. Here's the thing: AI IS NOT GENERALLY INTELLIGENT, AND WE CAN RELY ON IT TO FAIL AT 'LAST MILE' JUDGMENT WORK THAT REQUIRES GENERAL INTELLIGENCE. The goal here shouldn't be to remove humans from care. It should be to stop wasting clinical talent on work that doesn't require it. As routine tasks get automated, clinicians should have more room for nuanced conversation, the shared decision, the patient who needs persuasion as much as a prescription. Utah and NYU deserve real credit for running these experiments in the open.

27

6 Comments

I published a new piece with Silicon Foundry on a pattern that keeps repeating in AI infrastructure. Market leaders rarely miss these shifts because they’re asleep at the wheel. They miss them because they’re doing what leaders are trained to do. They run the playbook that made them leaders. They optimize the system they already have. And they trust the signals that have always correlated with strength. The problem is that AI infrastructure is not just more of the old stack. When architectures move, the basis of competition moves with them. Categories get redefined quietly, then all at once. By the time the change shows up cleanly in the numbers, the advantage has already migrated to a different layer of the value chain. In the article, I walk through why this happens and what to do about it. The ability to notice when the stack is reorganizing, before your planning process forces the answer back into last year’s categories. If you’re operating in semis, networking, or data center infrastructure, I’d be interested in how you’re watching for these shifts in your own system. https://lnkd.in/gimKJPUu

16

3 Comments

You can have groundbreaking science, elegant engineering, and even early customer interest. But without this one thing, you don't have an investable business. After two decades running IP strategy projects and now leading UC Berkeley's Deep Tech Innovation Lab while building the Berkeley Gateway Accelerator, I've watched countless brilliant technologies die in the valley between breakthrough and business. The problem isn't what most technical founders think it is. It's not about having better tech. It's not about getting more funding. It's not even about finding product-market fit. The fatal flaw shows up much earlier—and it's almost always the same mistake. In my latest piece, I answer questions from a Taiwanese entrepreneur about what really separates deep tech ventures that scale from those that stall. Including one provocative suggestion for Asian ecosystems that has nothing to do with technology. Read the full conversation, linked below. #DeepTech #Innovation #Startups #IntellectualProperty #VentureCapital

60

6 Comments

This week, Anthropic launched Claude for Life Sciences, its first real push to make AI part of how science gets done. Why this matters 👇 - Scientists spend hours each week buried in docs and data. Claude aims to give them that time back by summarizing literature, drafting protocols, and analyzing data. And with prebuilt connectors to Benchling, BioRender, and PubMed, it fits well into existing workflows. - Some researchers have been eager to use AI but hesitant to apply it without clear benchmarks. Claude’s most capable model, Sonnet 4.5, shows strong performance in lab protocol and bioinformatics tasks, making AI adoption in the field more credible and comfortable. - Anthropic also introduced Agent Skills, instructions that dictate how Claude tackles specific tasks. The first skill they’ve built performs single-cell RNA sequencing quality control using scverse best practices, a sign they’re prioritizing technical rigor. It's a strong step forward, but the rollout points to where pharma still is on the AI adoption curve. Instead of the self serve model offered to consumers, Anthropic is partnering with consulting firms like Deloitte, Accenture and KPMG to bring Claude to enterprise pharma. I can't help but wonder, if the goal is to meet scientists where they are, why would it take a consulting team to get started? 🤔 WDYT? Image Credit: Anthropic Agent Skills, announcement link in comments

87

3 Comments