sábado, 16 de maio de 2026

Quais as universidades e politécnicos que mais contribuíram para que Portugal esteja agora ao nível da produção da Suécia na área das publicações científicas sobre IA ?



Relativamente ao ranking que divulguei no post acessível no link supra, onde Portugal aparece exactamente com o mesmo rácio da Suécia, reproduzo abaixo a lista com a produção científica actualizada das universidades e politécnicos públicos. A análise das universidades e politécnicos que mais produziram relativamente ao ranking de Maio de 2025 é apresentada abaixo dessa lista. 

U.Porto..........................317 publicações indexadas na Scopus 
U.Lisboa........................254 
U.Minho........................188 
U.Aveiro........................170 
U.Coimbra.....................151 
U.Nova..........................150 
Polit. do Porto...............120 
ISCTE...........................117 
UTAD............................110 
UBI.................................80 
Polit. de Bragança..........49 
Polit. de Coimbra............43 
Polit. do Cávado e Ave....38 
Polit. de Leiria.................35 
U.Algarve........................33 
Polit. de Viseu.................21 
U.Évora...........................21 
Polit. de Portalegre.........20 
Polit. de Setúbal..............19 
Polit. de Lisboa...............19 
Polit. Viana do Castelo...18 
U.Aberta.........................17 
U.Madeira.......................13 
Polit. da Guarda..............12 
Polit. de Santarém...........11
Polit. de Tomar.................8
Polit. de Castelo Branco...6 
U.Açores..........................2
Polit. de Beja....................2 

A Universidade do Porto mantém a liderança nacional e é também quem mais cresce em número absoluto. Entre os politécnicos, o destaque maior vai para o Politécnico do Porto, que quase triplica a sua produção anterior. Uma evolução igualmente impressionante é a da Universidade de Aveiro que cresce 161,5% subindo dois lugares no ranking nacional. Em termos relativos, sobressaem também o Politécnico da Guarda, o Politécnico de Viseu, a Universidade de Évora, a Universidade Aberta e o Politécnico de Coimbra, vide Tabela 3. 

Tabela 1 - Principais subidas no ranking nacional

InstituiçãoRanking 2025 Ranking 2026 Variação
Politécnico de Viseu24.º16.º+8 lugares
Universidade de Évora22.º17.º+5 lugares
Politécnico da Guarda29.º24.º+5 lugares
Politécnico de Coimbra16.º12.º+4 lugares
Universidade de Aveiro6.º4.º+2 lugares
Instituto Politécnico do Porto9.º7.º+2 lugares


Tabela 2 - Maiores subidas em termos absolutos

Instituição2025 2026 Aumento  Crescimento
Universidade do Porto169317+148+87,6%
Universidade de Lisboa126254+128+101,6%
Universidade do Minho82188+106+129,3%
Universidade de Aveiro65170+105+161,5%
Universidade Nova de Lisboa69150+81+117,4%
Instituto Politécnico do Porto41120+79+192,7%
Universidade de Coimbra82151+69+84,1%
ISCTE54117+63+116,7%
UTAD47110+63+134,0%
Universidade da Beira Interior3680+44+122,2%


Tabela 3 - Maiores crescimentos em termos percentuais

Instituição2025 2026 Aumento Crescimento
Politécnico da Guarda212+10+500,0%
Politécnico de Viseu521+16+320,0%
Universidade de Évora621+15+250,0%
Universidade Aberta517+12+240,0%
Politécnico de Coimbra1343+30+230,8%
Instituto Politécnico do Porto41120+79+192,7%
Politécnico do Cávado e do Ave1338+25+192,3%
Universidade de Aveiro65170+105+161,5%
Universidade do Algarve1333+20+153,8%
Politécnico de Portalegre820+12+150,0%
à(s)

quinta-feira, 14 de maio de 2026

What Is Happening with Germany? AI Research Output Tells a Troubling Story

 


Around six months ago, in the post linked above, Germany’s Continued Underperformance and Portugal’s Escalating AI Obsession, I examined the roots of Portugal’s unusually high academic productivity in artificial intelligence. One result stood out in particular: according to a Scopus-based comparison, Portugal was already producing about 65% more AI-related publications per million inhabitants than Germany.

A similar search carried out today suggests that Germany’s relative position has become even more concerning. Several much smaller countries now show far stronger AI publication intensity than Germany. The gap is particularly striking in the cases of Cyprus and the United Arab Emirates. Compared with Germany’s AI publications per million inhabitants, Cyprus has an advantage of around 237%, and the UAE around 165%, over Germany.

These differences are too large to ignore. They raise an uncomfortable question about Germany’s ability to keep pace in one of the most strategically important scientific fields of this century. The institutional comparison is also revealing. Harvard Medical School alone reports 1,151 Scopus-indexed AI publications, while Germany’s leading university in this search, Technische Universität München, reports 707. The citation gap is also worrying: Germany has only 5 AI-related publications with more than 500 citations, compared with 10 in India, 11 in Canada, 15 in Australia, 18 in the United Kingdom and 33 in the United States.

For a country long associated with engineering excellence, scientific strength, industrial leadership, and technological sophistication, these numbers are difficult to dismiss. So the question must be asked: What is happening with Germany? Is Germany underinvesting in AI research? Is its academic system too slow, too fragmented, or too bureaucratic? Is talent moving elsewhere? Or are smaller and more agile countries simply adapting faster to the AI revolution?

Scopus AI-related publications per million inhabitants
1Singapore388 
2Cyprus 381
3United Arab Emirates299
4Ireland271
5Switzerland268
6Qatar262
7Finland259
8Norway248
9Bahrain244
10Australia209
11Greece208
12Denmark208
13Sweden192
14Portugal192
15Kuwait186
16Austria173
17Netherlands172
18Saudi Arabia169
19United Kingdom168
20Oman167
21Jordan155
22Canada140
23Italy140
24New Zealand126
25Croatia124
26Belgium116
27Germany113

Update after 1 day - Blogger analytics indicate that the majority of views for this post come from the USA (21%), Germany (12%) and Brazil (10%). 
à(s)

quarta-feira, 13 de maio de 2026

The Case of a Disease That Never Existed: How Fake Research Entered AI Systems

 

Almira Thunström, a physician at the University of Gothenburg, designed an experiment as elegant as it was unsettling. She invented an entirely fictitious medical condition—bixonimania, supposedly a skin disorder affecting the eyelids and linked to blue-light exposure—and uploaded two fabricated preprints to an academic repository. Within weeks, major large language models were presenting the condition to users as established medical fact. 

The experiment was deliberately constructed to be transparently fraudulent. The fictional lead author, Lazljiv Izgubljenovic, was listed as affiliated with the non-existent Asteria Horizon University in the equally imaginary Nova City, California. The acknowledgements thanked “Professor Maria Bohm at The Starfleet Academy” and cited funding from “the University of Fellowship of the Ring.” One paper explicitly stated that it had been entirely fabricated; another described its study population as “fifty made-up individuals.” Yet none of these obvious warning signs prevented the fabricated research from being absorbed into AI-generated medical knowledge. https://www.nature.com/articles/d41586-026-01100-y

Less than three weeks after the fabricated preprints appeared, major AI systems were already presenting bixonimania as a legitimate medical condition. Microsoft’s Copilot described it as “an intriguing and relatively rare condition.” Google’s Gemini advised users to consult an ophthalmologist. Perplexity AI even estimated its prevalence at one case per 90,000 people, while OpenAI’s ChatGPT offered symptom comparisons. The experiment suggests that the very architecture of academic publishing can become a vehicle for misinformation when AI systems are trained to treat these signals as proxies for credibility. More troubling still, the false research did not remain confined to AI-generated outputs. One of the fabricated preprints was later cited in a peer-reviewed article published in Cureus, part of the Springer Nature group, which described bixonimania as “an emerging form of periorbital melanosis linked to blue light exposure.” After inquiries from Nature, the journal retracted the paper on 30 March 2026—demonstrating how quickly fabricated research can move from preprint repositories to AI systems and, ultimately, into the scientific record itself.

PS - Yet the most obvious failure in this case was not the language models. It was the near-total absence of upstream verification. These papers were not sophisticated forgeries. They cited funding from “the University of Fellowship of the Ring,” listed an author affiliated with a university that did not exist, and located that institution in an imaginary city. And still, they passed through the submission process. Preprint repositories such as bioRxiv and medRxiv continue to accept submissions tied to institutions that may not exist, often with less scrutiny than a commercial web form. A basic automated check against established research registries such as the Global Research Identifier Database would have flagged “Asteria Horizon University” and “Nova City, California” in seconds. This is not a technical limitation, nor a question of cost. It is a question of institutional priorities. If platforms feeding the scientific record cannot detect references to fictional universities, imaginary cities, or funding bodies borrowed from fantasy literature, they should not be surprised when fabricated research flows downstream into AI systems, citation networks, and eventually the scientific literature itself.

à(s)
Subscrever: Mensagens (Atom)