Aquests dies de final d’estiu m’he trobat alguns articles sobre la intel·ligència artificial en química (i en ciència), que és interessant d’afegir als que ja he comentat en aquest blog. Per exemple, com recrear la taula periòdica dels elements, a partir de textos trobats a Internet: “Learning atoms for materials discovery“. De fet, phys.org deia que la taula periòdica s’ha recreats en poques hores gràcies a la AI, mentre que va tardar molt de temps a Mendeleiev de fer-la. Crec que això és una afirmació incorrecta. Fent servir tècniques d’anàlisi semàntica, a partir d’informació disponible a Internet, s’ha mirat quins elements van al cantó de quins altres elements. D’aquí s’ha generat una taula periòdica… senzillament, allò que a classe diríem que és una “anàlisi de clústers”. Si en Mendeleiev hagués tingut accés a la informació disponible actualment, amb uns bons ordinadors… segur que hagués construït la TPE ràpidament.
El Chemistry Worlk parla d’un laboratori química on la IA decideix què fer: Algorithm decides on chemical compromises when optimising self-driving experiments
Scientists in Canada have devised a machine-learning algorithm for optimising chemistry experiments. The algorithm, which they tested on a self-driving laboratory, judges the chemical merit of competing variables to identify the best conditions according to the user’s preferences.
El C&EN diu Machine learning could offer chemists fast, accurate calculations. Updated neural network system predicts molecular energies and forces faster than currently used methods (presentat a la Divisió de Computers in Chemistry de l’ACS, de la qual per cert en sóc membre):
Some researchers think machine learning could offer a better way. At the American Chemical Society national meeting in Boston on Tuesday, Adrian Roitberg of the University of Florida described a method that can achieve the accuracy of CCSD(T) in the computational time of force fields. He, along with Florida colleage Justin S. Smith and Olexandr Isayev of the University of North Carolina, Chapel Hill, call it Accurate NeurAl networK engINe for Molecular Energies (ANAKIN-ME).
During a session in the Division of Computers in Chemistry, Roitberg said the third version of the method, which the team calls ANI-1ccx, can predict the forces and energy of a molecule with only the positions of its atoms and their atomic number. The algorithm treats each element separately, then produces a summed prediction of the forces and energy in the molecule as a whole.
En Don Truhlar, químic reputat de la Univ. Minnesota, en diu el següent
If ANI-1ccx can continue to make relatively fast, accurate predictions, “there is a good chance it will become more useful than DFT for certain problems,” says Donald G. Truhlar of the University of Minnesota, who has worked on expanding the capabilities of DFT. He credits Roitberg for using a broad training set to avoid a common pitfall of machine learning algorithms, which often struggle when faced with molecules that look very different from those used in training. But he points out that ANI-1ccx may struggle with some situations that are difficult for all computational methods, like open-shell systems or excited states.
Em fa l’efecte que actualment cal per a alguns grups de recerca publicar qualsevol cosa i etiquetar-la amb “intel·ligència artificial”. Seguiré interessat en el tema, és clar…
El que està força bé és l’article crític de C&EN Is machine learning overhyped? Chemists weigh in on the technique’s possibilities and its pitfalls. Està bé el que s’hi diu: hi ha una cosa que el “machine learning” no és: màgia. M’agrada! 🙂
Machine learning is a category of artificial intelligence that describes a computer’s ability to train on a set of data and then create rules or knowledge from that data. Chemists are often interested in the tool’s predictive power. For instance, if you give a machine-learning algorithm a list of 100 metal alloys and their melting points, can it predict the melting point of an alloy it hasn’t encountered before—potentially even one that’s never been synthesized?
Despite all this promise—or perceived promise—one thing that machine learning isn’t is magic. “Let’s be realistic,” says George Dahl, a computer scientist at Google. “Machine learning is nonlinear regression,” a simple type of statistical analysis in which collected data are “fit” with model parameters. Dahl won a Merck & Co. machine-learning competition while a graduate student in Geoffrey Hinton’s group at the University of Toronto.
Una cosa queda clara: la IA és aquí i s’hi quedarà:
Despite their differences, the chemists that C&EN interviewed agree: Yes, machine learning is overhyped. No, it won’t cure cancer. Nonetheless, it’s a valuable tool that’s here to stay.
How valuable is up for debate. But in order for chemists to get the most out of machine learning, another thing is clear. Chemists need to change their behavior. That starts with data.
Una mica és un oxímoron, per no dir una contradicció:
Whether that’s overhyping it depends on your perspective. For those chemists who work most closely with machine learning, the excitement they see in press releases and casual conversation can get tiresome. These experts have great faith that machine learning will have a real and lasting impact on chemistry, especially if more people are trained to use it. At the same time, some worry that this tool can’t possibly live up to the highest expectations and that disappointment might hurt progress.
Cronin puts it this way: “Although I say machine learning is overhyped and annoying, I think it’s underused by chemists.”