Sign in

Spencer J Fox

@foxandtheflu.bsky.social
783 followers 378 following 192 posts

Assistant prof at NAU in disease modeling, and data science. Posts are my own opinions

PostsRepliesMedia
Spencer J Fox @foxandtheflu.bsky.social · 11/08/2026
We’re officially in Chennai to lead a workshop on forecasting and the MicroHub tool with the CDC and ICMR! 35 attendees from 5 countries 🤯
000
Spencer J Fox @foxandtheflu.bsky.social · 09/08/2026
Flying to Chennai to give our third forecasting workshop on the MicroHub tool we have been developing, and I am feeling grateful for all of the hard work my students + post-docs have put into our forecast efforts the past few years! Also grateful for free wifi on the plane 😱
000
Spencer J Fox @foxandtheflu.bsky.social · 24/07/2026
Finishing up the SISMID course on simulation-based inference of epidemiological models with instructions on how to run Bayesian, metapopulation, and phylodynamic models in pomp. Check out all of the lectures and exercises here: rubbislam.quarto.pub/episim/
020
Spencer J Fox @foxandtheflu.bsky.social · 23/07/2026
Our group traveled to Santiago with PAHO and the CDC to carry out a training on forecasting and our MicroHub tool, so that the MoH can run their own forecasts. PAHO just published a nice article about the visit if you're interested in the international work we're doing! www.paho.org/en/news/26-6...
paho.org
Chile implements respiratory virus forecasting with support from PAHO, CDC, and Northern Arizona University
Santiago, Chile, June 26, 2026 (PAHO) – Technical teams from the Chilean Ministry of Health, the Pan American Health Organization (PAHO/WHO), the U.S. Centers for Disease Control and Prevention (CDC),...
010
Spencer J Fox @foxandtheflu.bsky.social · 23/07/2026
2nd day of our simulation-baesd inference for epidemiological models course at SISMID! Qianying, teaching particle filtering approaches to estimate likelihoods in the pomp R package now! I always learn something new watching her teach this stuff 😅
000
Spencer J Fox @foxandtheflu.bsky.social · 22/07/2026
Such cool stuff, and I think I need to check under my couch cushions to get myself a swarm of agents for next flu season
000
Spencer J Fox @foxandtheflu.bsky.social · 22/07/2026
Really interesting talk at CSTE from Matt Doxey at google talking about using AI agents for public health modeling!
130
Spencer J Fox @foxandtheflu.bsky.social · 21/07/2026
Excited to be back in Atlanta for both the CSTE Forecasting workshop and SISMID! Hope to see some of y’all there
010
Spencer J Fox @foxandtheflu.bsky.social · 27/04/2026
America’s oldest public university, but technically UGA was the oldest public chartered university (go dawgs)
010
Reposted by Spencer J Fox
Joshua Weitz @joshuasweitz.bsky.social · 22/04/2026
Today, our #SCIMaP team released an impact analysis of the White House proposed FY 2027 NIH and NSF budgets. Bottom-line: if enacted, cuts to science and medical research would lead to $35B in economic losses and 150,000 lost jobs in communities all across the US. scienceimpacts.org 🧵
Landing page for scienceimpacts.org
1222151
Reposted by Spencer J Fox
Dr Roz Eggo @rozeggo.bsky.social · 12/04/2026
Great speakers at this symposium - 22 April 2026 at LSHTM. Join online or in person (email if in person) to hear great talks at the cutting edge of Inequalities in infectious disease dynamics. www.lshtm.ac.uk/newsevents/e...
lshtm.ac.uk
Inequalities in infectious disease dynamics symposium | LSHTM
Join us for the Social Inequalities in Infectious Disease Dynamics Symposium at the London School of Hygiene & Tropical Medicine (LSHTM). The day features a full programme of presentations and invited
11613
Spencer J Fox @foxandtheflu.bsky.social · 01/04/2026
I need this to be real, so I can get around NAUs rules against spending funds on coffee machines/accessories
020
Spencer J Fox @foxandtheflu.bsky.social · 31/03/2026
Based on my understanding of AI, I don't think having it "think" longer will address any of the issues I faced building the app, so I think there will need to be some major breakthroughs to get there (e.g. moving beyond LLMs or embedding them with new capabilities)
010
Spencer J Fox @foxandtheflu.bsky.social · 31/03/2026
I absolutely can see how AI will make coding more productive and even my research, but I don't see this replacing individuals for important tasks in the near-term and any use of this for research purposes should be heavily audited before trusting
100
Spencer J Fox @foxandtheflu.bsky.social · 31/03/2026
Maybe I'll come back to vibecoding the app again, but I am not sure. In the end I produced something that looked like it worked, but couldn't even make simple calculations accurately.
100
Spencer J Fox @foxandtheflu.bsky.social · 31/03/2026
This week I ended the experiment and finally opened up excel again... It was a breath of fresh air: I knew that the numbers I inputted would save properly, could fully audit calculations, and ultimately was able to get my budgets organized
100
Spencer J Fox @foxandtheflu.bsky.social · 31/03/2026
I thought I was making progress, the AIs were constantly finding new bugs! But then I tried to use it again and immediately I found new errors related to salary calculations and a number of things hidden within the app.
100
Spencer J Fox @foxandtheflu.bsky.social · 31/03/2026
All of this made me lose total confidence in the app, I wasn't going to spend a bunch of time inputting all of my real information unless I thought this would actually work. So I spent about a month working with the AIs to simplify the app and debug everything.
100
Spencer J Fox @foxandtheflu.bsky.social · 31/03/2026
The software was way outside of my understanding (Rust, SQL databases, etc), and I had no way to dig into what was going on under the hood. I would enter in numbers, save, reload and find that the numbers had changed
100
Spencer J Fox @foxandtheflu.bsky.social · 31/03/2026
Then about a month ago (and after I confirmed with both AIs that the app was ready for primetime) I started really using it. Immediately I came across issues, the app had lied to me about database structures, saving features, loading features, budget calculations, basically everything.
100
Spencer J Fox @foxandtheflu.bsky.social · 31/03/2026
On the good side, the app was built pretty quickly and it was really cool to see the interface and play around with the features. It was incredible to produce an app just for myself and I even started having hopes of getting it good enough I could post online for others in academics to use
100
Spencer J Fox @foxandtheflu.bsky.social · 31/03/2026
I found myself spending this vibe code time "coding" and also answering emails, which wasn't a terrible use of time, but I would have rather been writing or thinking.
100
Spencer J Fox @foxandtheflu.bsky.social · 31/03/2026
First, the experience is probably about as bad as it can get. Your agent works long enough that you have to try and work on other things while it's "thinking" but not long enough for you to get anything meaningful done in that time
100
Spencer J Fox @foxandtheflu.bsky.social · 31/03/2026
I probably spent about 2 months working on the app on and off, adding and removing features, testing that it was working, etc. Overall it was probably about 40 hours of work across that time period with both codex and claude editing the app and critiquing each others work
100
Spencer J Fox @foxandtheflu.bsky.social · 31/03/2026
This is something that I have a fairly simple excel spreadsheet to handle that I was hoping to move beyond, as I thought it would be more convenient, flexible, and ultimately accurate.
100
Spencer J Fox @foxandtheflu.bsky.social · 31/03/2026
Starting from scratch I worked with the AI to develop what really should not be hard to develop. An app that allows me to enter information about various grants, personnel costs, expenses, and then can track the spend so that I can know how much money I have and where it is, etc.
100
Spencer J Fox @foxandtheflu.bsky.social · 31/03/2026
To really test the AI I worked with it to vibe code a grant planning app, since budgeting various grant spend down and personnel salary costs within university environments is basically the bane of my existence (numbers and grants here are fake)
210
Spencer J Fox @foxandtheflu.bsky.social · 31/03/2026
Double checking for small and often really dumb errors can take longer than coding things up myself, so I actually found I was using the AI tools only for limited coding tasks (e.g. I code up the full pipeline and have the AI input code in specific areas to do specific tasks)
100
Spencer J Fox @foxandtheflu.bsky.social · 31/03/2026
As you get further towards full full agency, double checking becomes really important. When they are developing their own models, I needed to look through the code closely to identify both common and edge case errors that would have caused issues if I used for real or research purposes
100
Spencer J Fox @foxandtheflu.bsky.social · 31/03/2026
For example, I have pipelines for producing infectious disease forecasts weekly. I can confidently use these tools to modify the code to use the data in different ways, with slight model variants, and even some simple and totally new model variants that I have the code develop itself
100
Spencer J Fox @foxandtheflu.bsky.social · 31/03/2026
In my testing I found that these tools work really well when I have existing code that needs to be modified e.g. with aesthetic changes to tools, slight model modifications, new data inputs, data cleaning changes, etc. Even refactoring a whole code repository goes pretty smoothly
100
Spencer J Fox @foxandtheflu.bsky.social · 31/03/2026
I have been playing with Codex and Claude extensively for about 6 months now, and while they are clearly impressive, I don't see them replacing software engineers in the near future and I don't understand how people can trust their work without knowing how to code
110
Spencer J Fox @foxandtheflu.bsky.social · 27/03/2026
I'm here for the vibes! Would love to chat more :)
010
Spencer J Fox @foxandtheflu.bsky.social · 26/03/2026
Yea I'll take a closer look at the basket index! Ensembling baselines is an interesting idea, but what would the point be? Capturing different reference opinions or improving performance?
110
Reposted by Spencer J Fox
Sam Abbott @seabbs.bsky.social · 25/03/2026
I really like this idea of multiple baselines with different contexts. some relationship to this basket of baselines idea I was chatting about a while ago community.epinowcast.org/t/a-basket-o...
community.epinowcast.org
A basket of baselines
Been thinking about the overlap between baseline models (Manuel Stapper is working on this), the method of analogs (stuff @nickreich and co are going on about), and model output evaluation (i.e. @kat...
111
Spencer J Fox @foxandtheflu.bsky.social · 24/03/2026
Congratulations to Fox Lab PhD student @ehsansuez.bsky.social for his hard work on this project!
020
Spencer J Fox @foxandtheflu.bsky.social · 24/03/2026
Our recommendation: Transparency and use multiple Baseline models: "...include an optimal flatline model as a stringent benchmark, a robust flatline model, and a seasonal baseline to test whether information from the current season improves predictions."
220
Spencer J Fox @foxandtheflu.bsky.social · 24/03/2026
Why this matters: Forecast hubs rely heavily on baselines for: • benchmarking models • ranking performance 👉 If the baseline shifts, the leaderboard can change too. Our work is complementary and supports the conclusions from previous work: www.medrxiv.org/content/10.1...
medrxiv.org
Mind the Baseline: The Hidden Impact of Reference Model Selection on Forecast Assessment
Baseline models are essential reference points for evaluating forecasting methods, yet their selection often receives insufficient attention. We present a systematic framework for baseline model selec...
110
Spencer J Fox @foxandtheflu.bsky.social · 24/03/2026
Interestingly, this flatline model outperformed seasonal models even with highly seasonal data
110
Spencer J Fox @foxandtheflu.bsky.social · 24/03/2026
Our main finding: Performance differed substantially across variations AND a flatline model that uses the ten most recent transformed observations for training outcompeted all others across all diseases
110
Spencer J Fox @foxandtheflu.bsky.social · 24/03/2026
Baseline models are often treated as “neutral” benchmarks. But they involve subjective choices. We evaluated how three choices impact the performance: 1. The training window 2. Data transformations 3. Whether forecasts have slopes (drift) or are constant (flat)
110
Spencer J Fox @foxandtheflu.bsky.social · 24/03/2026
🚨 New preprint from our lab! How much does the design of “simple” baseline models matter for forecast hubs? We show that small design choices in baseline models can substantially change model performance, and thus the relative rankings of the other models www.medrxiv.org/content/10.6...
medrxiv.org
Basic Baseline model design choices can substantially influence performance in collaborative forecast hubs
Over the past decade, outbreak forecasting has become an increasingly used tool to assist public health decision-making during epidemics. Collaborative forecast hubs, where multiple teams submit predi...
193
Spencer J Fox @foxandtheflu.bsky.social · 26/02/2026
Attending the @gs-idd.bsky.social launch seminar. Couldn't be more excited to see this effort take shape and can't wait to see how it helps push the infectious disease dynamics community forward! Thank you to all of the founding members and organizers 🙏
022
Reposted by Spencer J Fox
Elizabeth Jacobs, PhD @elizabethjacobs.bsky.social · 24/02/2026
Republicans in Arizona want voters to get rid of childhood vaccine requirements for school entry. They are planning to add a ballot proposition for November’s election in the hopes of ending the requirements, which they are framing as an issue of “medical freedom”. This would be catastrophic.
azcentral.com
Arizona voters could decide future of government vaccine mandates
The Arizona Legislature is considering whether to refer a proposed constitutional amendment to the 2026 statewide general election ballot.
2418588
Reposted by Spencer J Fox
Cevian Labs @cevianlabs.io · 24/02/2026
Today we’re excited to announce a new product: COI Pond is a tool for faculty and grant staff that converts an academic CV into a draft NSF Collaborators and Other Affiliations (COA) Form. COA forms are tedious work, and COI Pond makes it a snap. Try it for free, here cevianlabs.io/coipond/start/ /1
Screenshot of the COI Pond webpage: 

Upload your academic CV and generate a draft NSF Collaborators and Other Affiliations (COA) Form to submit with your NSF grant proposals.

COI Pond parses your CV to extract:
Coauthors of articles published, in the last 48 months
Collaborators on funded projects, in last 48 months
Past PhD advisees & your own PhD advisor(s)
Editorial roles, in last 24 months
After extraction, you can edit the draft COA form before downloading a final version (in xslx format) that you can submit to NSF.

COI Pond is a free tool* from Cevian Labs.
1145
Reposted by Spencer J Fox
Dan Larremore @danlarremore.bsky.social · 24/02/2026
You know that annoying NSF form "List every coauthor/co-PI from the last 4y" ? At @cevianlabs.io we built a free tool that drafts the COI form from your PDF CV in minutes. Check it out 👇
13513
Spencer J Fox @foxandtheflu.bsky.social · 23/02/2026
If every student uses it then your class discussion boards will just be AI talking to AI endlessly, slowly adding CO2 to the atmosphere...
000
Spencer J Fox @foxandtheflu.bsky.social · 23/02/2026
My prediction is it will be really obvious to tell if a student uses something like this (at least for a while).
100
Spencer J Fox @foxandtheflu.bsky.social · 23/02/2026
Who is this for? If you don't know the material from the class, how can you review the outputs? If you have it auto-submit why are you even paying for the class? companion.ai/einstein
companion.ai
Einstein - AI Homework Agent
Einstein logs into Canvas and does your homework automatically. He has his own computer — he can watch lectures, read essays, write papers, and participate in discussions.
100
Reposted by Spencer J Fox
Adam Kucharski @adamjkucharski.bsky.social · 23/02/2026
Interested in learning outbreak analytics and modelling, with a focus on real-life case studies and tools actively used in responses? Apply for the popular @lshtm.bsky.social R shortcourse, led by @jwlambert.bsky.social & @andreevallecam.bsky.social More: www.lshtm.ac.uk/study/course...
1179