Felipe R. da Silva.
LightNews LightNews
felipers.bsky.social
Felipe R. da Silva.
@felipers.bsky.social
240 followers 320 following 470 posts
Bioinformatics @Embrapa.
Posts Replies Media Videos
Reposted by Felipe R. da Silva.
andrew.heiss.phd
Some closing thoughts for my students this semester on LLMs and learning #rstats datavizf25.classes.andrewheiss.com/news/2025-12...
Will you incorporate LLMs and AI prompting into the course in the future? No. Why won’t you incorporate LLMs and AI prompting into the course? These tools are useful for coding (see this for my personal take on this). However, they’re only useful if you know what you’re doing first. If you skip the learning-the-process-of-writing-code step and just copy/paste output from ChatGPT, you will not learn. You cannot learn. You cannot improve. You will not understand the code. In that post, it warns that you cannot use it as a beginner: …to use Databot effectively and safely, you still need the skills of a data scientist: background and domain knowledge, data analysis expertise, and coding ability. There is no LLM-based shortcut to those skills. You cannot LLM your way into domain knowledge, data analysis expertise, or coding ability. The only way to gain domain knowledge, data analysis expertise, and coding ability is to struggle. To get errors. To google those errors. To look over the documentation. To copy/paste your own code and adapt it for different purposes. To explore messy datasets. To struggle to clean those datasets. To spend an hour looking for a missing comma. This isn’t a form of programming hazing, like “I had to walk to school uphill both ways in the snow and now you must too.” It’s the actual process of learning and growing and developing and improving. You’ve gotta struggle. This Tumblr post puts it well (it’s about art specifically, but it applies to coding and data analysis too): Contrary to popular belief the biggest beginner’s roadblock to art isn’t even technical skill it’s frustration tolerance, especially in the age of social media. It hurts and the frustration is endless but you must build the frustration tolerance equivalent to a roach’s capacity to survive a nuclear explosion. That’s how you build on the technical skill. Throw that “won’t even start because I’m afraid it won’t be perfect” shit out the window. Just do it. Just start. Good luck. (The original post has disappeared, but here’s a reblog.) It’s hard, but struggling is the only way to learn anything. You might not enjoy code as much as Williams does (or I do), but there’s still value in maintaining codings skills as you improve and learn more. You don’t want your skills to atrophy. As I discuss here, when I do use LLMs for coding-related tasks, I purposely throw as much friction into the process as possible: To avoid falling into over-reliance on LLM-assisted code help, I add as much friction into my workflow as possible. I only use GitHub Copilot and Claude in the browser, not through the chat sidebar in Positron or Visual Studio Code. I treat the code it generates like random answers from StackOverflow or blog posts and generally rewrite it completely. I disable the inline LLM-based auto complete in text editors. For routine tasks like generating {roxygen2} documentation scaffolding for functions, I use the {chores} package, which requires a bunch of pointing and clicking to use. Even though I use Positron, I purposely do not use either Positron Assistant or Databot. I have them disabled. So in the end, for pedagogical reasons, I don’t foresee me incorporating LLMs into this class. I’m pedagogically opposed to it. I’m facing all sorts of external pressure to do it, but I’m resisting. You’ve got to learn first.
December 9, 2025 at 8:17 PM
Reposted by Felipe R. da Silva.
alemartinello.com
Super!

One point you might make even more explicit, which is obvious for you but not always to the students, is that the product of your class is not whatever output the students produce. Nobody cares about it. It will be thrown away.

The product is learning.
December 9, 2025 at 10:52 PM
Reposted by Felipe R. da Silva.
revdjesq.bsky.social
Calibri was asking for it, walking around without any serifs.
December 11, 2025 at 12:30 PM
Reposted by Felipe R. da Silva.
merriam-webster.com
The first known use of ‘conspiracy theory’ was in 1863…or so they want you to believe.
December 8, 2025 at 9:48 PM
Reposted by Felipe R. da Silva.
benjaminjriley.bsky.social
I'll stop here. But note that these are all cases involving ADULTS outsourcing the basic requirements of their professional responsibilities to these tools of cognitive automation.

Now tell me we need to be pushing this into schools. I fucking dare you.
December 4, 2025 at 2:24 PM
Reposted by Felipe R. da Silva.
atila.bsky.social
Colocaram implantes cerebrais em uma mulher que perdeu os movimentos e agora ela consegue tocar piano só com os impulsos nervosos enquanto ela imagina tocar. Muito legal.

O twist: os eletrodos lêem os impulsos antes de serem registrados pela consciência. Então ela sente que o piano toca sozinho.
Mind-reading devices can now predict preconscious thoughts: is it time to worry?
Ethicists say AI-powered advances will threaten the privacy and autonomy of people who use neurotechnology.
www.nature.com
December 2, 2025 at 5:17 PM
Reposted by Felipe R. da Silva.
meganranney.bsky.social
Gift article about the across-the-board *slowdown* in federal funding for science - on ALL the topics. But especially for up-and-coming scientists.

www.nytimes.com/interactive/... 🎁
The U.S. Is Funding Fewer Grants in Every Area of Science and Medicine (Gift Article)
A quiet policy change means the government is making fewer bets on long-term science.
www.nytimes.com
December 2, 2025 at 2:02 PM
felipers.bsky.social
“it is easier to imagine the end of the world than the end of capitalism.”
December 2, 2025 at 2:44 PM
Reposted by Felipe R. da Silva.
economist.com
Three years into the generative-AI wave, demand for the technology seems surprisingly flimsy
Investors expect AI use to soar. That’s not happening
Recent surveys point to flatlining business adoption
econ.st
November 29, 2025 at 9:20 PM
felipers.bsky.social
"Meanwhile, artificial intelligence technologies are undermining the value of the industry’s creative endeavours by enabling companies to create ads quicker and more cheaply".
December 1, 2025 at 2:59 PM
Reposted by Felipe R. da Silva.
lindaholmes.bsky.social
Here’s this year’s Books We Love—but thanks to lots of work by lots of people, you can also look over all the Books We Love years back to 2013. It’s a remarkable project, and I can say that because I have nothing to do with it. (/fin)
Books We Love
Here are 380+ great reads from 2025 handpicked just for you by NPR staffers and trusted critics.
apps.npr.org
November 30, 2025 at 1:54 PM
Reposted by Felipe R. da Silva.
ars.deichator.ar
it’s over
hetanshah.bsky.social Hetan Shah @hetanshah.bsky.social · 14d
Your universe is a photocopy of a photocopy of a…
on.ft.com/4ahkNKa
Al-generated content has surpassed human-written articles Graph
November 28, 2025 at 10:49 AM
Reposted by Felipe R. da Silva.
mahideia.bsky.social
dá pra ter muito problema ao mesmo tempo não precisa escoher um só.
November 26, 2025 at 6:21 PM
Reposted by Felipe R. da Silva.
charles.littlegreenfootballs.com
i’m not crying you’re crying

xkcd: Fifteen Years

Fifteen Years
xkcd.com
November 25, 2025 at 1:58 AM
Reposted by Felipe R. da Silva.
arstechnica.com
Many genes associated with dog behavior influence human personalities, too
But the specific behaviors linked may be completely unrelated.
arstechnica.com
November 26, 2025 at 12:33 PM
Reposted by Felipe R. da Silva.
davelevitan.bsky.social
Google at its peak was basically the best information retrieval system in human history and they and every competitor decided going from there to “you didn’t want answers you wanted half-assed auto-complete 80%-wrong hallucinations” in a few years was the right idea
November 25, 2025 at 1:57 AM
Reposted by Felipe R. da Silva.
kamounlab.bsky.social
I just published: Gene Editing: Precision Breeding Regulations Now Live

On 13 November 2025, UK regulations for precision bred plants came into effect. We can now register precision bred plants, such as those with disease resistance and reduced reliance on fungicides.

medium.com/p/gene-editi...
Gene Editing: Precision Breeding Regulations Now Live
FIRST PUBLISHED: 14.11.25 BY LAURA TURCHI AND MIA CERFONTEYN | THE SAINSBURY LABORATORY
medium.com
November 25, 2025 at 2:21 PM
Reposted by Felipe R. da Silva.
msoares.bsky.social
Salve, pessoal. Um projeto que eu estava preparando para o ano que vem subiu no telhado. Cheguei a recusar algumas propostas pra não estar com o tempo comprometido caso desse certo. Não deu. Assim, estou aberto a tomar cafés pra pensar coisas pra adiante.
November 24, 2025 at 6:16 PM
felipers.bsky.social
Chorei no início de uma aula na semana retrasada falando do Lô. Enquanto eu viver, garanto a imortalidade dele!
November 24, 2025 at 5:00 PM
Reposted by Felipe R. da Silva.
adamgrant.bsky.social
It's time to remove laptops from classrooms.

24 experiments: Students learn more and get better grades after taking notes by hand than typing. It's not just because they're less distracted—writing enables deeper processing and more images.

The pen is mightier than the keyboard.
November 20, 2025 at 3:09 PM
Reposted by Felipe R. da Silva.
benjaminjriley.bsky.social
Although focused on the EU, the warnings offered here are no less true for AI policy in the US, or anywhere else for that matter. In my lifetime I've never seen grandiose hype of what *might* happen with a technology drive what we are doing in the here and now. Not good!
November 17, 2025 at 2:35 PM
Reposted by Felipe R. da Silva.
benjaminjriley.bsky.social
It's a wee bit unfortunate that the phrase "cognitive offloading" has become the go-to phrase for describing what AI does, including by those worried about its impact on education. I believe a more accurate term is cognitive *automation*, as Francois Chollet described years ago.
AI is cognitive automation, not cognitive autonomy
Like the rest of computer science, AI is about making computers do more, not replacing humans.
fchollet.substack.com
November 17, 2025 at 3:05 PM
Reposted by Felipe R. da Silva.
icij.org
ICIJ's latest investigation The #CoinLaundry is a collaboration of 113 journalists from 38 media partners in 35 countries that exposes how cryptocurrency companies have empowered a shadow economy that lavishly profits from crime.

Here are our findings:
November 17, 2025 at 3:08 PM
felipers.bsky.social
Even @wired.com acknowledges how superior (better!) @home-assistant.io is!

I've been using Home Assistant for almost 2 years now. There's no turning back.
November 16, 2025 at 3:46 PM