Weeknote WN202022 – sumant

Demà es passa a la nostra regió sanitària a la Fase 2 de la desescalada del confinament per la Covid-19. En la pràctica no varien gaire les coses… excepte que si fa calor ja es pot anar a la platja. I les escoles i centres de secundària tornen a obrir parcialment (a Girona ciutat dimarts, perquè demà és Cinquagesma) amb dubtes, molts de dubtes. A aquesta petita millora hi he de sumar el fet que se’ns ha acceptat una comunicació oral (en línia, per descomptat) al congrés EDEN 2020 de Timisoara, a finals de juny.

El símbol d’aquests dies és la mascareta (una altra d’aquestes contradiccions en què vivim). M’ha agradat veure un cub de rubik de 12x12x12 amb la imatge de la mascareta:

Aquests dies tenia previstos de passar-los a Aberdeen (Escòcia), al congrés PCST 2020. Hi teníem una comunicació llarga. Però s’ha ajornat a l’any vinent. Llàstima, perquè no he estat mai a Escòcia, i el tema de la trobada és molt interessant (Public Communication of Science and Technology).

Aquesta setmana he participat en dues activitats rellevants en línia, que ja he recollit al blog: dimarts, el congrés de Suma d’Intel·ligències, organitzat per la UdG, i ahir la 3a sessió del #cafèdigital16, liderat per la @mferres. La casualitat (o no?) ha fet que avui hagi aparegut al Diari Ara un reportatge a doble pàgina sobre la docència virtual i la visió que en tenen els estudiants joves, amb un paper destacat de la Titi Juher, professorat de literatura i amb qui coincideixo en diverses iniciatives: Com ho han fet els mestres? Cal obrir les escoles? Responen els alumnes

La imatge de la setmana

M’ha semblat rellevat aquest acudit sobre els estudiants (americans) que es graduen al mig de la pandèmia de #covid19…

El que he vist per la xarxa

Portem molts de dies en confinament i post-confinament, i es parla poc de notícies no relacionades amb la Covid-19. A When Will It Be OK to Push Non-COVID-19 News? s’hi comenta com, en comunicació interna i externa, les universitats han de començar a donar notícies no relacionades amb la pandèmia, però mirant de fer-ho de forma positiva.

With internal audiences, there is much more room for pushing these positive stories — and the first wave of shifting back to some sense of normalcy with news will be within campus networks. Institutions need to continue fundraising efforts and maintain enrollment, and these stories play a role in meeting those critical institutional goals. We can — and should — be celebrating our recent graduates and student achievements, profiling campus and community members who are doing extraordinary things in these trying times and highlighting the work and accomplishments of professors. We need to re-examine campus newsroom functions and internal audience reception as it becomes more critical to shift to forward-looking news and opportunities for students.


El J. Med. Chem. té un article titulat Chemists: AI is here, unite to get the benefits  que analitza la relación entre la química mèdica i la química computacional, en particular com amb les sòlides tècniques QSAR es poden fer moltes coses. Ja ve bé que es critiqui constructivament la AI.

The latest developments in artificial intelligence(AI) have arrived into an existing state of creative tension between computational and medicinal chemists. At their most productive, medicinal and computational chemists have made significant progress in delivering new therapeutic agents into the clinic. However, the relationship between these communities has the prospect of being weakened by application of over-simplistic AI methods which, if they fail to deliver, will reinforce unproductive prejudices. We review what can be learnt from our history of integrating QSAR and structure-based methods into drug discovery. Now with synthesis and testing available as contract services, the environment for computational innovation has changed and we consider the impact this may have on the relationships in our disciplines. We discuss the current state of interdisciplinary communication and suggest approaches to bring the sub-disciplines together in order to improve computational medicinal chemistry and, most importantly, deliver better medicines to the clinic faster.

Precisament aquesta setman a Slashdot s’hi comentava com, de fet, no hi ha hagut avenços reals en AI darrerament: Eye-Catching Advances in Some AI Fields Are Not Real  comenta l’article de Science  amb le mateix títol

Artificial intelligence (AI) just seems to get smarter and smarter. Each iPhone learns your face, voice, and habits better than the last, and the threats AI poses to privacy and jobs continue to grow. The surge reflects faster chips, more data, and better algorithms. But some of the improvement comes from tweaks rather than the core innovations their inventors claim — and some of the gains may not exist at all, says Davis Blalock, a computer science graduate student at the Massachusetts Institute of Technology (MIT). Blalock and his colleagues compared dozens of approaches to improving neural networks — software architectures that loosely mimic the brain. “Fifty papers in,” he says, “it became clear that it wasn’t obvious what the state of the art even was.” The researchers evaluated 81 pruning algorithms, programs that make neural networks more efficient by trimming unneeded connections. All claimed superiority in slightly different ways. But they were rarely compared properly — and when the researchers tried to evaluate them side by side, there was no clear evidence of performance improvements over a 10-year period. The result [PDF], presented in March at the Machine Learning and Systems conference, surprised Blalock’s Ph.D. adviser, MIT computer scientist John Guttag, who says the uneven comparisons themselves may explain the stagnation. “It’s the old saw, right?” Guttag said. “If you can’t measure something, it’s hard to make it better.”

Researchers are waking up to the signs of shaky progress across many subfields of AI. A 2019 meta-analysis of information retrieval algorithms used in search engines concluded the “high-water mark … was actually set in 2009.” Another study in 2019 reproduced seven neural network recommendation systems, of the kind used by media streaming services. It found that six failed to outperform much simpler, nonneural algorithms developed years before, when the earlier techniques were fine-tuned, revealing “phantom progress” in the field. In another paper posted on arXiv in March, Kevin Musgrave, a computer scientist at Cornell University, took a look at loss functions, the part of an algorithm that mathematically specifies its objective. Musgrave compared a dozen of them on equal footing, in a task involving image retrieval, and found that, contrary to their developers’ claims, accuracy had not improved since 2006. “There’s always been these waves of hype,” Musgrave says.

I de química…

Una mica relacionat amb l’article esmentat sobre QSAR, m’ha fet gràcie que ha estat un meu homònim parcial, en Miquel Duran Frigola, del BIST, qui ha publicat Database offers another way to search for drugs – by activity not structure. Es tracta de trobar compostos ja existents que puguin tenir noves aplicacions per a les quals ningú no hi ha pensat encara. Felicitats!

A new chemical dataset provides an alternative way for biomedical researchers to search for promising compounds. The Chemical Checker contains bioactivity data on almost one million molecules and has already been used to identify substances with previously unknown medicinal traits.

I de comunicació científica…

L’article Meeting in the middle del Chemistry World parla de què, passada la pandèmia i tornats a una situació de nova normalitat, potser podem quedar-nos amb algunes de les coses bones de fer trobades en línia:

What I’ve found most valuable is the chance to hear from speakers anywhere in the world without extra costs. I recently watched a fantastic, pan-European drug discovery symposium delivered via weblink. The feedback I saw was very positive, and it was exactly as engaging as if the speakers were in the room.

It looks like I’ll be running a couple of online meetings soon, a new experience for most attendees and organisers alike. We are excited to widen accessibility, include delegates from more countries and add enhanced technology such as auto-captioning and improved live-tweeting. Just like for an in-person meeting, a vast amount of work goes into the audience experience every time. Organisers should make the most of this current period of self-reflection, and strive to retain opportunities for increased accessibility and international inclusion when we can once again meet in person.

Els meus tweets de la setmana