Sign in

Adolfo Neto

@adolfoneto.elixiremfoco.com
5.7K followers 4.6K following 17K posts

Software Engineering researcher; Associate Professor at UTFPR, Curitiba, Brazil. adolfoneto.elixiremfoco.com Posts mostly in English. #ElixirLang enthusiast (I also love #LeanLang, #LuaLang, #Erlang). Podcaster.

PostsRepliesMedia
Adolfo Neto @adolfoneto.elixiremfoco.com · 8h
Zoom: Tommorrow
Lampedusa Keeps Happening: Refugees and the Power of Tech to Harm or Heal 

Date & Time 	Oct 1, 2026 06:00 PM Amsterdam, Berlin, Rome, Stockholm, Vienna 
Webinar ID 	812 8479 9762 
Description 	On October 3, 2013, 368 people, mostly Eritrean refugees, lost their lives in the Lampedusa shipwreck. On the 13th anniversary of the tragedy, join DAIR’s "Refugees, Migrants, and AI" project for an event that commemorates those who died and asks why, more than a decade later, Lampedusa keeps happening.

Bringing together refugee advocates and experts in international law, psychology, and forced migration, the conversation will connect the human consequences of contemporary border regimes with the growing use of surveillance, data and AI in migration control. We will ask what accountability looks like. What would it take for technology to protect people on the move rather than deepen the risks they already face?

The event will include: 

- Keynote by Meron Estefanos, Researcher at DAIR, journalist, and activist who played a key part in documenting the identities of all the victims of the Lampedusa shipwreck.

Panel and Q&A with:
- Dr. Omer Shatz — Lecturer in International Law, Sciences Po Paris; Co-founder and Co-director, front-LEX
- Dr. Selam Kidane — Lecturer in Psychology & Systemic Psychotherapist, University of West London
- Hyab Teklehaimanot Yohannes — Lecturer in Forced Migration & Decolonial Education, UNESCO Chair, School of Education, University of Glasgow
010
Adolfo Neto @adolfoneto.elixiremfoco.com · 12h
"Don't believe self-serving lies about technologies being "inevitable" or "here to stay". You don't have to just go along with the dominant narrative. You can make deliberate choices and help others to do the same." ozark.hendrix.edu/~yorgey/fore...
Brent Yorgey
153
Adolfo Neto @adolfoneto.elixiremfoco.com · 13h
Poste algo aleatório, ou você terá um péssimo Outubro!
Imagem da gravação do Emílias Podcast com Nina da Hora
040
Adolfo Neto @adolfoneto.elixiremfoco.com · 29/09/2026
Yet another new model Is it really different from the previous one?
A screenshot introduction a new LLM model from a big ai company
280
Adolfo Neto @adolfoneto.elixiremfoco.com · 29/09/2026
"It is an AI revolution"
A funny robot in an OpenAI livestream
150
Adolfo Neto @adolfoneto.elixiremfoco.com · 28/09/2026
A colleague who wrote a book on Pre-Calculus now specializes in making AI-generated videos for Instagram. The videos have some math in them, but I refuse to watch them.
AI-generated images from videos that include cats and math
150
Adolfo Neto @adolfoneto.elixiremfoco.com · 28/09/2026
DHH: "I have unlimited tokens. We can fix everything, we can do anything." "dhh needs his meds" Lain on the Blockchain (CryptoCyberia on X)
A screenshot of a forum.

Open issue: Add an x86-64 assembly engine for all 37 effects 
#35
dhh wants to merge 167 commits into master from asm-zen5

nonetrix commented:
No offence, but why make this less portable for likely zero perceivable speedup? You could literally go frame by frame and probably spot nothing reliably on hardware made in the last 30 years, margin of error if anything at all. The only realistic difference between this and the old version is that it doesn't work on ARM or whatever other architecture now, a check to see if this version could be used likely has more overhead. This is 101 definition of misplaced priorities in my opinion
👍 6

dhh commented: "The pursuit of excellence requires no justification". And if you'd look at the code, you'd see that we fall back to the slower Rust implementation when asm acceleration is not possible.Also, you're still stuck in fixed-pie thinking if you think this is happening at the expense of anything else. I have unlimited tokens. We can fix everything, we can do anything.
😁 2 🎉 2 🚀 5
3132
Adolfo Neto @adolfoneto.elixiremfoco.com · 26/09/2026
Ótimo comentário sobre Cangaço Novo


gabriela.iris.t 2d
@carvalhoalice Cangaço assim como A roda do tempo tinha mulheres na produção e transmitia ao público uma necessidade urgente de sororidade, e união de classe. Cangaço trata também de grilagem para mineração, colocando a união da comunidade como possibilidade de saída. Como principais acionistas da Amazon (proprietária do Prime Video) insiders e gestores investem individualmente em empresas mineradoras de terras raras. Cabaceiras, Seridó, locais de gravação da série são potenciais locais para mineração de terras raras. Estamos em ano eleitoral e muito se fala de interferência norte-americana nas eleições, acredito que eles (acionistas da Amazon) já fizeram suas apostas no Brasil.
042
Adolfo Neto @adolfoneto.elixiremfoco.com · 24/09/2026
And the worst thing is that these reviews are bad.
Paul Ralph wrote:

Got your #ICSE #reviews? Gobsmacked by the parade of superficial, incompetent reviewing? You are not alone. It is not your imagination. Many of the reviews are, objectively, garbage. This is not just because of AI. Many of the reviews have always been garbage because evaluating papers by writing essays is fundamentally stupid and ineffective. Quality control systems use standards and checklists, like these: https://lnkd.in/eRDaEtkn
040
Adolfo Neto @adolfoneto.elixiremfoco.com · 23/09/2026
Is this a nightmare?
A sample assignment, for example, might begin with asking the student to brainstorm about a specified topic on a particular text—say, Toni 
Morrison’s Beloved—with the goal of producing a prose paragraph that 
can be submitted to ChatGPT as a prompt, asking the LLM to construct 
an outline from the paragraph.7
 Since AI responses are sensitive to the 
precise working of a prompt, including its tone, prompt construction can 
be effectively addressed in a workshop situation, where all the students 
have an opportunity to see what works and what doesn’t. When students 
fashion their own prompts, the student’s ideas should provide the basic 
material, while the AI may help to order them effectively (often a problem for developing writers). In addition, ChatGPT may also extend the 
student’s ideas by offering suggestions of its own. These texts, including 
prompt and response(s), would be included as part of the process material, 
along with the student’s critical evaluation of ChatGPT’s suggestions (“this 
idea strikes me as really good, the other one is too broad to be useful”). The 
student should also be encouraged to devise more than one prompt, and/
or to submit the same prompt multiple times, and to evaluate which of the 
program’s responses will likely be the most useful—again annotating and 
submitting them as part of the process package. The next step can be the 
student’s editing the program’s outline, refining its topics and developing 
the ideas further, for example with specific textual examples. The essay’s 
first draft should be constructed by the student, which can then be submitted to the AI for refinement, eloquence, grammar corrections, and so 
forth. The student should then edit the AI’s suggestions, explaining in a 
supplementary text (perhaps annotations on the refined draft) why the student made the indicated corrections and additions. These process materials would then be submitted along with (or perhaps before) the final essay
291
Adolfo Neto @adolfoneto.elixiremfoco.com · 23/09/2026
So Mythos is not as good as they said. Surprising fosstodon.org/@KernelRecip...
Mythos's 79 vulnerabilities

24 - no detail at all "something crashed"
14 - not a bug at all
3 - totally made up data
15 - already fixed in latest release
  11 by others
  4 by Anthropic
070
Adolfo Neto @adolfoneto.elixiremfoco.com · 22/09/2026
I looked for The Nerd Reich, by @gilduran.com, on Amazon Brazil and found, in addition to his book: - El Reich de Los Nerds: El fascismo de Silicon Valley y la guerra contra la democracia, Gulle Deon - The Exit Plan Nerd Reic(...), Irene Mendez Navarro
Amazon screenchot:

The Nerd Reich, Gil Durán

- El Reich de Los Nerds: El fascismo de Silicon Valley y la guerra contra la democracia, Gulle Deon

- The Exit Plan Nerd Reic: Tech Oligarchy, the Network State, and Silicon Valley's Escape from Democracy, Irene Mendez Navarro
130
Adolfo Neto @adolfoneto.elixiremfoco.com · 21/09/2026
Abstract, according to Google Scholar
[CITAÇÃO] Don't Ban AI from Your Writing Classroom; Require It!
NK Hayles
Poetics Today, 2024•read.dukeupress.edu
The release of OpenAI’s ChatGPT is a game changer for college and university teachers worried about students using the remarkable linguistic abilities of large language models (LLMs) to pass off artificial intelligence (AI) writing as their own work. Of particular concern is the college essay, widely used in the humanities to assess student comprehension of complex texts, as well as to enhance their ability to create original, well-structured, and rhetorically effective written arguments. Stephen Marche’s (2022) proclamation “The College Essay is Dead” typifies the concern. 1 Others have rushed to defend the college essay (Rim 2022; Downey 2022; Roose 2023), with some professors arguing that AIs such as ChatGPT can only write pedestrian essays that might, at best, earn some kind of B grade, but not reach into the A range (Stacey 2022). From my point of view, these arguments miss the larger issues at stake in the natural language processing abilities of ChatGPT and similar LLMs. I suggest that their development should be seen as an opportunity to rethink the basic principles behind prohibitions against plagiarism and uses of the college essay to facilitate
read.dukeupress.edu
110
Adolfo Neto @adolfoneto.elixiremfoco.com · 21/09/2026
N. Katherine Hayles says LLMs are hideously innefficent. www.instagram.com/reels/Ddeogb...
47525
Adolfo Neto @adolfoneto.elixiremfoco.com · 19/09/2026
Qual é a próxima fronteira da Engenharia de Software? Graziela Tonin (INSPER) responde
150
Adolfo Neto @adolfoneto.elixiremfoco.com · 18/09/2026
News article or ad? > Claude will have 'Word' and slides inside the chat and change logic
Claude terá 'Word' e slides dentro do chat e muda lógica
020
Adolfo Neto @adolfoneto.elixiremfoco.com · 18/09/2026
Ghost in the Machine is probably being imdb-bombed I bet those 19 people that gave it a 1 never watched the movie. www.imdb.com/pt/title/tt3...
imdb ratings graph with 22 10 and 19 1
394
Adolfo Neto @adolfoneto.elixiremfoco.com · 18/09/2026
This is the Dario we truly respect
Dario, Dadá Maravilha, Soccer World Cup champion in 1970
140
Adolfo Neto @adolfoneto.elixiremfoco.com · 18/09/2026
AUTOMATIZEI
5H DE TRABALHO POR DIA

ENTÃO VOCÊ
VAI TRABALHAR
MENOS AGORA, NÉ?

VOCÊ VAI TRABALHAR
MENOS, NÃO VAI?
0312
Adolfo Neto @adolfoneto.elixiremfoco.com · 16/09/2026
James Zou • 2nd

Associate Professor at Stanford University

4h • New posts

Following

Thrilled that #paper2agent is published in Nature today! Papers have been static documents for centuries. Paper2Agent turns them into active AI agents that can answer questions, apply their methods, and collaborate with other papers to make new discoveries.Paper: https://lnkd.in/gf3cisYc
Code: https://lnkd.in/g79tdYRe
Paper2Agent agent: https://lnkd.in/gVk7j-p4Papers are written for humans. Simply giving a PDF to LLM can miss key insights. We want the virtual author to gain hands-on experience w/ the work not just read it. Our key idea: let the AI set up a virtual lab to run code, test methods and reproduce the paper's results.The result is an automatically generated paper mcp that captures the paper's key know-hows, tools and data. Giving a LLM this paper mcp is much more accurate and efficient than giving it the original paper + code repo!Agents of different papers can also collaborate with each other! For example, the AlphaGenome agent worked with a GWAS dataset agent to characterize splicing mutations associated with ADHD risk.Agentifying knowledge makes research easier to reproduce, reuse and extend.Great job by Jiacheng Miao with fantastic collaborators Jonathan Pritchard, Joe Davis and Yaohui Zhang 👏
220
Adolfo Neto @adolfoneto.elixiremfoco.com · 16/09/2026
Homenagem a Cecilia Rubira no @cbsoft.bsky.social
030
Adolfo Neto @adolfoneto.elixiremfoco.com · 15/09/2026
Corte de Inteligência Artificial na Saúde: Motivações e Tecnologias, Dra. Cíntia Baulé (DoctorAssistant.ai) www.youtube.com/watch?v=ChCv...
010
Adolfo Neto @adolfoneto.elixiremfoco.com · 15/09/2026
Addding Againstversary to my calendar
110
Adolfo Neto @adolfoneto.elixiremfoco.com · 14/09/2026
Sou oficialmente Professor Titular I am officially a Full Professor sei.utfpr.edu.br/sei/controla...
Boletim de Serviço Eletrônico em 11/09/2026


Ministério da Educação

​UNIVERSIDADE TECNOLÓGICA FEDERAL DO PARANÁ

Portaria de Pessoal GABIR/UTFPR nº 1504, de 11 de setembro de 2026

 

O REITOR DA UNIVERSIDADE TECNOLÓGICA FEDERAL DO PARANÁ (UTFPR), considerando o Decreto datado de 20 de setembro de 2024, publicado no D.O.U. de 23 subsequente;

considerando o contido no Processo nº 23064.012028/2026-98;

considerando o art. 12 da Lei nº 12.772, de 28 de dezembro de 2012, na redação dada pela Lei nº 15.141, de 2 de junho de 2025; e

considerando o Regulamento da UTFPR para Promoção de Docentes da Carreira de Magistério Superior à Classe de Professor Titular aprovado pela Deliberação nº 014, de 22 de agosto de 2014, do Conselho Universitário,

R E S O L V E

conceder PROMOÇÃO PARA A CLASSE TITULAR, a partir de 24/03/2026, a ADOLFO GUSTAVO SERRA SECA NETO, matrícula SIAPE n.º 1142878, ocupante do cargo efetivo de Professor do Magistério Superior, lotado no Campus Curitiba, conforme regulamenta a legislação supramencionada.

 

 

PUBLIQUE-SE E REGISTRE-SE

​

Gabinete da Reitoria
25862
Adolfo Neto @adolfoneto.elixiremfoco.com · 13/09/2026
Even though I think I'll be deeply disappointed with the rest of the interview, I enjoyed the beginning of this video with Terence Tao
320
Adolfo Neto @adolfoneto.elixiremfoco.com · 11/09/2026
Evandro Costa (UFAL) fala sobre agentes @cbsoft.bsky.social
0101
Adolfo Neto @adolfoneto.elixiremfoco.com · 11/09/2026
Podcaster desconhecido encontra podcaster famoso (Sérgio Amadeu, do Tecnopolítica)
Sergio Amadeu e Adolfo Neto
3360
Adolfo Neto @adolfoneto.elixiremfoco.com · 11/09/2026
Yesterday I met @ramalho.org and we talked about, among other things, AI in software engineering
Luciano Ramalho and Adolfo Neto
3271
Adolfo Neto @adolfoneto.elixiremfoco.com · 10/09/2026
The Association for Human Mathematics (AHM) (...) In light of the grave societal, environmental, and economic impacts of the AI industry and the threat of unfettered AI growth, we strive not to be complicit.
The Association for Human Mathematics (AHM)
The Association for Human Mathematics (AHM) protects mathematics as a human endeavor against the threat of artificial intelligence. We work to maintain the independence of the mathematical community from the encroachment of corporate interest. In light of the grave societal, environmental, and economic impacts of the AI industry and the threat of unfettered AI growth, we strive not to be complicit.
1195
Adolfo Neto @adolfoneto.elixiremfoco.com · 09/09/2026
University of São Paulo
0130
Adolfo Neto @adolfoneto.elixiremfoco.com · 09/09/2026
Excellent talk by Klaas Jan-Stol at @cbsoft.bsky.social Why are you doing this study? On the role of theory in software engineering research cbsoft.sbc.org.br/2026/pt/symp... You can read his paper: dl.acm.org/doi/10.1145/...
Klaas Jan-Stol
000
Adolfo Neto @adolfoneto.elixiremfoco.com · 06/09/2026
Sorry, Adam Conover, but @anthonymoser.com is my favorite AI hater
1264
Adolfo Neto @adolfoneto.elixiremfoco.com · 02/09/2026
XP 2026 no Brasil: um relato sobre uma das mais importantes conferências de métodos ágeis, com Dra. Graziela Simone Tonin (INSPER) fronteirases.github.io/episodios/pa...
Foto de Graziela Tonin em arte do Fronteiras
021
Adolfo Neto @adolfoneto.elixiremfoco.com · 01/09/2026
Nova tendência do mercado de moda no Brasil: "modelos" feitos por IA
0112
Adolfo Neto @adolfoneto.elixiremfoco.com · 01/09/2026
What happened to Lingvanex here? It translated "Whataboutism" to "E quanto ao socialismo" (And as for socialism). I couldn't find a good translation for Whataboutism, but Google Translate gave me: 'E quanto ao "e o que dizer do..."' (And what about "and what about...")
Original in English:
Whataboutism. This is an accusation of hypocrisy that is used to brush off concerns about the externalities of these systems. You eat meat, you fly on airplanes, etc, etc, so how dare you talk about the impacts of data centers? This is a recipe for inaction, on any front.

Lingvanex translation:
E quanto ao socialismo. Esta é uma acusação de hipocrisia usada para ignorar preocupações sobre as externalidades desses sistemas. Você come carne, voa em aviões, etc, etc, então como ousa falar sobre os impactos dos data centers? Esta é uma receita para a inação, em qualquer frente.
220
Adolfo Neto @adolfoneto.elixiremfoco.com · 31/08/2026
'IA' na Engenharia de Software (ES) é a 'técnica' mais ineficiente da história da ES? adolfoneto.elixiremfoco.com/blog/posts/2...
Sobre a ineficiência da "IA" na engenharia de software
1151
Adolfo Neto @adolfoneto.elixiremfoco.com · 30/08/2026
"WE ESPECIALLY POINTED OUT TO VEGETARIANS THAT THEY DID WEAR LEATHER SHOES"
The world is on fire but also full of water.
You can see some houses half under water.
A kid and a couple (man+woman) are on a boat. 

The kid: BUT WHAT DID YOU DO WHEN SOMETHING COULD STILL BE DONE ABOUT THIS?


The woman: WE ESPECIALLY POINTED OUT TO VEGETARIANS THAT THEY DID WEAR LEATHER SHOES


Original text:

MAAR WAT DEDEN JULLIE DAN TOEN HIER NOG IETS AAN GEDAAN KON WORDEN?


WE WEZEN VOORAL VEGETARIËRS EROP DAT ZE WEL LEREN SCHOENEN DROEGE
151
Adolfo Neto @adolfoneto.elixiremfoco.com · 30/08/2026
Why young people hate AI even more?
Lincoln Michel 
Counter Craft

Subscribe button

The industry is incapable of any self-reflection, but it should be obvious why this would be the case. Older gens saw Big Tech innovations that changed people's daily lives for the better and/or were at least mixed bags. Personal computers, internet, mp3 players, smart phones, etc. Someone who is 18 has seen.... what? NFT scams and crypto rug pulls. Web 3.0 hype that went nowhere. Gig economy apps that exploit workers and still charge too much for poor service. Zuckerberg's metaverse. A bunch of crap that massively enriched some billionaires without seeming to improve people's lives. Why should any young person look at big tech companies hyping GenAI—who were all pushing metaverse/crypto/etc. a few years ago—and think it would help them?
2143281295
Adolfo Neto @adolfoneto.elixiremfoco.com · 29/08/2026
Yes, I am going to @cbsoft.bsky.social @se4fp.github.io I don't like to travel that much, but CBSOFT this year is very special. It will happen at IME-USP, where I did my Doctorate.
An image containing my photo and information, in Portuguese, saying that I am going to CBSOFT 2026 and that I am SE4FP organizer and author of SE4FP and SBES papers
070
Adolfo Neto @adolfoneto.elixiremfoco.com · 29/08/2026
Is this a golf tee?
Blue object with "Blue Devils" written on itTop part of the blue object
210
Adolfo Neto @adolfoneto.elixiremfoco.com · 27/08/2026
Trecho
292
Adolfo Neto @adolfoneto.elixiremfoco.com · 27/08/2026
It's funny that one of the most inefficient technologies of all time (LLMs) is being used to help people write papers in software engineering that, as Jeff wrote, "has a very explicit goal of improving the way we develop software--better, cheaper, or faster"
Linkedin post by  Jeff Offutt (Professor and Chair at University at Albany)

Solve problems, don't count beans.

My research community is scared about a recent extreme increase in paper submissions to conferences. Many of them are low quality with little or no research contribution. And the most alarming trend is the number of paper submitted by the same author ... 10, 20, 50!

If we allow our names to be put on a paper, we should have read it and made a MEANINGFUL research contribution to the research. And being the money-bringer is not a research contribution. Even at my best, I couldn’t make a real contribution to more than 6 papers. Anyone who sends 15 or 20 or 40 papers to the same conference isn’t doing research. At best it’s fraud. 

What is our goal as a researcher? Are we truly only playing a numbers game, where more bullets on our CVs is a measure of our worth? How about refocusing away from numbers, and try to make contributions. New knowledge, better solutions to actual problems, increased understanding.

Most of my research is in software engineering, which has a very explicit goal of improving the way we develop software--better, cheaper, or faster. We write science fiction (it could work in the future), not fantasy (assumes magic or different fundamental laws of the universe). If the contributions in a paper have no chance to eventually help real software engineers, then why publish? Why do the work?
43913
Adolfo Neto @adolfoneto.elixiremfoco.com · 27/08/2026
#TT Horseshoe bend Page, Arizona July 2014
Horseshoe bend
020
Adolfo Neto @adolfoneto.elixiremfoco.com · 27/08/2026
UTFPR Ecoville
A lawn and some trees
040
Adolfo Neto @adolfoneto.elixiremfoco.com · 27/08/2026
Karen Hao on AI=Colonialism www.instagram.com/reel/DbqBIeN...
1135
Adolfo Neto @adolfoneto.elixiremfoco.com · 26/08/2026
Curiosidade: 13 anos atrás aconteceu a defesa de dois alunos orientados por mim, Maria Claudia Emer e Laudelino Cordeiro Bastos: Ramiro Batista Da Luz e Marcelo Fermann Guimaraes Alguns dias depois, mais dois: Marcos Antonio Chiarello e Vagner Carlos Marcolino Lima


Nome do aluno, data de defesa, Primeiro nome orientador, segundo coorientador(a)

Ramiro Batista Da Luz	26/08/2013	Adolfo Gustavo Serra Seca Neto	Robinson Vida Noronha
Marcelo Fermann Guimaraes	26/08/2013	Laudelino Cordeiro Bastos	Maria Claudia Figueiredo Pereira Emer

Marcos Antonio Chiarello	29/08/2013	Maria Claudia Figueiredo Pereira Emer	Adolfo Gustavo Serra Seca Neto
Vagner Carlos Marcolino Lima	27/08/2013	Adolfo Gustavo Serra Seca Neto	Maria Claudia Figueiredo Pereira Emer
000
Adolfo Neto @adolfoneto.elixiremfoco.com · 25/08/2026
Can we use LLMs in a good way???
132
Adolfo Neto @adolfoneto.elixiremfoco.com · 24/08/2026
Soon at @fronteirases.bsky.social: @lidianycs.bsky.social and I interviewed Simone Eringfeld, author of Podcasting as a Research Method
Simone, Adolfo and Lidiany
032
Adolfo Neto @adolfoneto.elixiremfoco.com · 24/08/2026
Agora sim! No dia 27 de agosto, às 19h, acontece o aulão online e gratuito "As Big Techs e o extrativismo verde", com a economista política e pesquisadora @sabrinafernandes.bsky.social (Instituto Alameda) e a jornalista investigativa Naira Hofmeister (The Guardian, Mongabay, El País).
Aulão
As Big Techs e o Extrativismo Verde
27/08
19h00
Com Sabrina Fernandes e Naira Hofmeister

Nijna
Pulitzer Center
14115
Adolfo Neto @adolfoneto.elixiremfoco.com · 23/08/2026
AI slop vs Design @sandsvisual on Instagram www.instagram.com/reel/DcJ6eQW...
0103