I’m looking for an automated way to read others’s scientific data without giving credit or acknowledgement, and also claim full credit for insights from it. And I want it to have a fitting name
OAI: say no more
👀 "U.S. STRIKES KHARG ISLAND: Even before Trump’s deadline, an intense wave of strikes was reported on bridges across Iran and on Kharg Island, the country’s key oil export hub."
Do you find it difficult to understand why LLMs — like ChatGPT and the like — are a form of ‘automated plagiarism’?
Take 3 min and listen to this excellent explanation by @gwenvarley.bsky.social
youtu.be/1ON-FJHq9aM?... (h/t @olivia.science) 1/🧵
Excited to share our new Prosit-PTM paper is out! A deep learning model that predicts MS2 spectra and Retention time for modified peptides, using chemically informed encodings and data augmentation based on amino acid substitution to learn general rules for how PTMs affect spectra.
Looking for a tool to estimate abundances of aerobes versus anaerobes directly from metagenomes? Here you go: www.biorxiv.org/content/10.6...
📢 Job opening — Bioinformatics Engineer
Join the MicroScope platform (LABGeM, CEA/Genoscope) to develop tools and pipelines for microbial genomics.
📍 Evry, France | ⏳ Fixed-term contract
🔗 bit.ly/3Z5WSXc
@ifb-elixir-fr.bsky.social
👀 New blog post! 👀
Ever made a *spaghetti* line chart and wondered how you can make it more accessible and more aesthetically pleasing at the same time? 📈
Read this blog post: nrennie.rbind.io/blog/accessi...
#DataViz #RStats
Thanks for writing and re-upping this. I think terminology is important, but I don't think regulating based on technical or statistical properties (e.g., "generative" or "transformer") is appropriate. I think the mode of interaction and epistemic function is more important for regulation.
Grading and googling hallucinated citations, as one does nowadays, and now that LLMs have been around for a while, I've discovered new horrors: hallucinated journals are now appearing in Google Scholar with dozens of citations bc so many people are citing these fake things
NEW: My Udemy course 'Propaganda & Disinformation for Beginners' has helped students in 26 countries worldwide strengthen their armour against disinformation since it launched 2 years ago. Techniques Cambridge Analytica used to secretly profile millions have become a vastly profitable industry.. /1
Who says capitalism can't innovate when OpenAI is out here inventing consumer liability protection law from first principles
“There were no cervical cancer cases diagnosed in women aged <25 years in 2021 – the first time that this has occurred in data going back to 1982. This remarkable achievement is almost certainly due to the impact of HPV vaccination.” From report.cervicalcancercontrol.org.au
#vaccinationsaveslives
More people need to read "Why does he do that?" because a great deal of people are bullied and we don't turn into abusers. The narrative these people are victims is perverse to the goal of stopping abuse because it's not only false but contributes the cycle.
dn790007.ca.archive.org/0/items/Lund...
Exited to share our latest work! Out now in @natcomms.nature.com
Koina aims to transform how #proteomics uses machine learning. You no longer need to be a tech wizard to use ML and now can easily run #ML models. Integrated with FragPipe, Skyline and EncyclopeDIA!
www.nature.com/articles/s41...
GcMeta - a new global resource of metagenome-assembled genomes and their encoded functions with an easy to use, interactive and searchable website academic.oup.com/nar/advance-...
The amount of AI generated art in slides at this conference, primarily used by older scientists, is killing me. Scientists please. Don’t use these ai platforms to make your figures or slides. They look bad and I have yet to see them meaningfully improve the message of talks.
A reminder.
Our new review paper about Pathway-Centric PTM Data Analysis is out this week in #Proteomics!
We cover databases, enrichment tools, software for pathway reconstruction, and full-fledged platforms that help to interpret high-throughput PTM datasets.
Check it out here: doi.org/10.1002/pmic...
One throughline in all these AI promises is eliminating subjectivity. We don’t have to trust people’s accounts of their own experiences, because we can trust a magic machine’s *estimates* of those experiences. let ChatGPT write your memoir, act as “synthetic” interview respondents, choose yr dates..
For the avoidance of doubt, this is how the inventor of "vibe coding" actually writes code.
Can we all stop pretending now?
🧠 The Lipid #Brain Atlas is out now! If you think #lipids are boring and membranes are all the same, prepare to be surprised. Led by @lucafusarbassini.bsky.social with Giovanni D'Angelo's lab, we mapped membrane lipids in the mouse brain at high resolution.
www.biorxiv.org/cgi/content/...
Bacterial telomeres are common, just not so much in RefSeq 'complete' genomes. But they can be added by the new tool David Faurdal wrote. I am thrilled to see this out as a preprint here: www.biorxiv.org/content/10.1... @tilmweber.bsky.social @thombooth.bsky.social
One way of thinking about the AI bubble is that it's the final gasp of a grift based on specialized GPU's instead of tulips, a logical continuation from crypto and NFT's and the metaverse.
Like, there's nowhere else for it to go, no other magical use for this shit other than as electronic waste.
"Our final defenses are more diffuse, working at a level of norms and attitudes. Stigmatization is a powerful force, and disgust and shame are among our greatest tools. Put plainly, you should feel bad for using AI." (finally reading this)
We have multiple senior scientist / engineer positions available across several geos in MSR AI for Science. Ping me or any of my colleagues if you have any questions!
www.microsoft.com/en-us/resear...
Preprint and the website by @dingdingpeng.the100.ci and @vincentab.bsky.social are great, and I got inspired to illustrate Bayesian workflow for model checking and comparison *before* model interpretation following their friendship importance example users.aalto.fi/~ave/casestu...
