Greg Medlock

@bcherny He's been cloned, beaming up: bsky.app/profile/did:pl… @bcherny-x.bsky.social

Hi @bcherny can you mirror this on bsky? Should pretty much be one-shot. Claude code dev discussion is basically my only value derived from X…

@shelbynewsad OG self-amplifying drugs: live biotherapeutic products & fecal microbiota products

(new thesis) self-amplifying drugs Healthy humans are a wonder. Their bodies sense danger (viruses, cancer, bacteria), kill danger, develop from fetus to adult,  repair wounds, and so much more. Healthy bodies should be the standard for what drugs of the future look like. Drugs should sense, respond, heal, regrow, and do this more and more cheaply over time. Human bodies are diagnostic and therapeutic and our drugs eventually should be too - e.g. cell therapies, tissues therapies, organ replacement. In the journey to this future there are improvements that can be made to our current modalities to make them look closer to living systems. A core step closer to living, breathing drugs is self-replication. Eg. how you make the drug produce more of itself in humans? In this vein, many papers have come out in the past years  specifically on [self-replicating  mRNA therapies](sciencedirect.com/science/articl… COVID-19 pandemic resulted,been approved against COVID-19.) (saRNA). Whereas mRNA vaccines only encode one COVID spike protein, saRNA vaccines encode for RNA replication machinery + a COVID spike protein, enabling the RNA to replicate itself in your cells. What this means is that a materially small amount of saRNA (5 μg) can have the same effect as mRNA at 30-100 μg. Incredibly, there’s already an approved COVID saRNA vaccine in Japan and Europe that uses this technology, called Kostaive. The company that developed it is a small cap biotech, Arcturus ($ARCT) but has had limited uptake ($28M revenues for Kostaive in 2024, company decreased revenues significantly in 2025). Kostaive also has similar level of side effects. There’s other preclinical work on saRNA on heart attack repair, Zika, Ebola, HIV, and Toxoplasma gondii. Tissue repair seems to be a good, acute use-cases from wounds, to spinal cord injury to bone regeneration there acute factors that could be secreted to continuously dose during this period. saRNA in vivo cell therapy could add durability to the response of CAR-T or saRNA could allow for better delivery of gene therapy. Other Thoughts The RNA self-replication is a hallmark of viruses which our bodies are primed to detect and destroy. saRNA drugs are destroyed in a matter of 4 weeks with very high immune responses to them. Different delivery vehicles are proposed as a way to limit initial immune response. But repeat dosing won't be as effective and destroyed much faster. Eg. how many acute interventions are possible to treat. What makes this modality 10x better for normal diseases? For vaccines there’s more immunity. Why aren’t all biologics of the future dosed via mRNA? Second order effects of less spend or plateau’d spend on manufacturing. Comparable Companies Replicate Bio Strand Bio Amplitude Arcturus Therapeutics Gennova Biopharmaceuticals HDT Bio VLP Therapeutics Chimeron Bio Keylicon Biosciences

@ibddoctor @michelioe Peter has been hacked

Despite continued resistance to preprints from some academics/editors, plenty of people DNGAF where your work is published Data point for trainees: if I am reviewing your job app or interviewing you, I actually read your papers.

Press release: businesswire.com/news/home/2026…

Vedanta Biosciences Announces the Phase 3 RESTORATiVE303 Study of VE303 for the Prevention of Recurrent C. difficile Infection Will Continue as Planned Following Protocol-Specified Interim Analysis

Onwards! Vedanta Biosciences Announces the Phase 3 RESTORATiVE303 Study of VE303 for the Prevention of Recurrent C. difficile Infection Will Continue as Planned Following Protocol-Specified Interim Analysis

@UrbanCourtyard @BBillDavis @PhilSustainable I had never seen this broken down, and couldn’t find anything more recent than this ‘06 inventory. Has this been pretty stable since ‘06 or has the distribution changed?

@BBillDavis @PhilSustainable Stuy town is majority studios and 1BDs

America's high-cost areas need ~4M more homes. To fill that gap, we need to build 356 Stuyvesant Towns (11,250 units each). All market rate. Doing this would slash housing costs, boost local economies, reduce homelessness, & cut emissions. It's time to build at scale

@DrSynbio Within gene tx, what are some examples of CMC requirements that are too stringent? Why are they too early?

Regulatory reform means right sizing/staging CMC requirements and also funding early/small scale infrastructural build outs to accelerate new ideas safely getting into patients. Especially for genetic medicines, this would be an exponential and accelerating change.

All comes down to COL. Basically every other place in USA has the same issues highlighted in article (if anything, Boston/Cambridge are more resilient due to being such a concentrated hub for our health/adjacent sectors). Can’t be solved through incentives internal to tech or other sectors. Need residents to stop pulling up the ladder behind them, e.g., fighting the MBTA communities act.

A pretty rough article came out today re the Massachusetts economy (see in reply). While I agree with a lot of the problems, a lot of the builders in Massachusetts are working very hard to make it not just a great place to start companies, but also a great place to scale companies. Recent progress: 1. YC - There are a lot of folks helping fog the runway for a YC return to Massachusetts. In @agupta we trust. 2. DeltaV - This is MIT's incubator that got a $6million infusion of cash and programming reboot from @abialecki and @edhallen of @klaviyo fame. 3. TechWeek - This is going to be a major event in Massachusetts this year. Big boost happening. 4. Sequoia - Running programming in Massachusetts. Thanks to @abhishekm1636 5. Mass AI Coalition - A major initiative across all of tech in the state led by @willahmed and @durkin 6. Government - The mayor, governor, @epaley and @Barry_Finegold are starting to face reality and are engaging in the problems. Green shoots include a statewide partnership with OpenAI and a citywide initiative to teach AI in all the schools. I'm really hoping to see real "action" from our representatives on building housing and improving transportation. 7. MIT - MIT's president and provost have kicked off their CATE CATE (Committee for Acceleration Translation and Entrepreneurship) initiative that I'm hopeful will have a big impact. I'm sure I'm leaving out a ton more. What am I missing @lindahenry, @alexwg, @patk, @HenryLSchuck, @mikey, @BillAulet?

@UrbanCourtyard @criticalurban He’s a troll. Blind corners are awful when managing more than 1 kid under 5, and are only getting worse with the cybertruckification of sight lines.

Good for you! That is also my experience with one kid. I’m talking about parents with multiple children, and about parents who wish to allow their kids to free range in the neighborhood without worrying they will get smashed at an intersection There have been two child deaths in my neighborhood since we moved here. Both were caused when a very young child suddenly ran into an intersection with bad visibility (driver couldn’t see them)

I emphasize chamfered corners in building designs because I've been TRAUMATIZED by years of parenting little kids in a city with too many blind corners. 90 degree corners mean you can't SEE what is approaching the intersection and may hit you (or your little kid who is small and not neurotically scanning the roads for cars like you are). POV you have a kid in stroller, you're walking a dog, and you are shouting "MIND THE ALLEY" to older kid zooming ahead on the bike.

@tallinzen (But i agree—the back bay stop should not exist)

@tallinzen It is bizarre because there are only two stops in Boston

Really bizarre that the Acela makes three stops in Boston, but only one in NYC, a much bigger city

@WalentekLab @SGRodriques @JocelynnPearl @EdisonSci Anecdotally: not at all. These tools have only been net-useful for ~6 months. Not long enough for organizations to adapt.

@MedlockGreg @SGRodriques @JocelynnPearl @EdisonSci C'mon industry already oversubscribed to the whole thing.

Welp, I guess this is the path... of LLMs? My favorite LLM for scientific research (FutureHouse's Platform) is now part of @EdisonSci and only offers a $200/month option or enterprise option. Nothing in between? Sad because I was actually using it to help patients.

Good reminders from @mitsuhiko. Not just for software dev. Cannot count the number of times the “roots” have anchored me: “I or someone else, planted something, and then I kept showing up, and eventually the thing had roots that went deeper than my enthusiasm on any given day.”

@WalentekLab @SGRodriques @JocelynnPearl @EdisonSci Academic users are future industry subscribers

@SGRodriques @JocelynnPearl @EdisonSci Nothing is free. So are you taking the researcher work to train your model or sell their user stats/info or both?

Agree with sentiment, but the logic can be disproven by reviewing your own paper from >1-5 years ago. Knowledge, judgement, taste change within individual. Under some circumstances, an individual provides better “external” feedback to themselves than the available (and willing) alternative sources.

Remember that you, by definition, cannot distinguish between a good idea and a failure of your own knowledge, judgement, or taste. This is why external (ideally expert) feedback is crucial

@bobbyfijan Same shock moving Charlottesville -> Cambridge during infant days. On weekends most Cambridge coffee shops open at 8AM, vs plentiful 7AM options in small college town… Dunkin is a drain on the universe

Living in Center City as a parent of young kids ... the most ironically annoying thing about the coffee shops, was that not even Starbucks opened before 7am. Dunkin was, literally, the only option.

@bijans Auchinsloss? PBMs arguably bigger source of waste than healthcare fraud.

Some of this healthcare fraud video stuff is of course performative and bad use of data. But a lot of it isn't. Yet, not a single Dem that I'm aware of at national/state/big city level is making the case for good government, efficiency, and catching and stopping healthcare fraud at its core. They've handed the entire narrative to Rs. That reputational hole / letting voters brand Ds as sclerotic, wasteful spenders, beholden to interest groups and who enrich fraudsters and raise your taxes to do so vs. the “efficient” R image will take a generation to climb out of.