Our experimentations on AI have induced destruction of our main server biatri-1, purchased in 2015 via the BIATRI project of Prof. Bulgakova and managed by T. J.-Y. Derrien.
Learning by doing. Learning by doing mistakes, too.
Our experimentations on AI have induced destruction of our main server biatri-1, purchased in 2015 via the BIATRI project of Prof. Bulgakova and managed by T. J.-Y. Derrien.
Learning by doing. Learning by doing mistakes, too.
The world has transformed into a completely real-time connected entity. ChatGPT and Claude.ai have transformed the world already, into something that I may have dreamed about being young. Somehow, it seems I anticipated it would happen but had kept producing publications of high quality at a moderate pace in an environment that was craving for quantity.
It seems AI can enable to have both: the quality and the quantity. And the quantum too! 🙂
As a result, it became difficult to manage priorities. Because everything is important. AI can help to tackle this priority, but privately-owned AI software may surely have their own priorities. That makes things quite complicated to manage. Putting your relationships at risk. Increasing loneliness. Reducing standards sometimes.
More details can be provided upon unreasonable request.
Our Octopus-slabs code, that T. J.-Y. Derrien has been developing since 2016 is working better than if asking a code-specific AI engine to run the Octopus computations by itself.
Basically, the AI engine submits jobs out of control of the job management system and can literally destroy the machine by running too big computations.
Alessio started today with us and will stay for a few months using the Erasmus+ program.
We have then more competence to fight against safety issues, now.
All the best,
TJY Derrien
Our latest paper on modeling 2D materials on substrate using quantum modeling has been submitted. It is available in ResearchGate for the moment. More details to come.
For 1 year, I have been spending a large amount of time using LLM for multiple topics (psychology, mathematics, physics). The experience I have made with this tool is frankly intense.
Hence, using AI today appears to be both stupid and clever. For those who believe that the quickest gets it all in this world, they will run for it by themselves! For others who depend on relationships and on human warmth, they will look their way around from this, and take the risk to be forgotten in few years. So as usual, taking the middle may enable to get the best of this revolution.
Meanwhile, high schools teach their students to program few Python lines, whereas these skills are already fully automated on machines, at very low financial cost for the end user – but at very high environmental cost.
In that context, energy production is foreseen to keep increasing. Making the world more sustainable will be happening mostly for those who are ready to make sacrifices. Others will simply keep high consumption – which is not proportional to results. Governments will take the decision of whom will have to make sacrifices and some resistance will appear in some parts of the world.
To whom tomorrow’s world will be driven by, then? To those who will search for their inner thoughts and will and nourish their surrounding with it? To those who will run the quickest with the latest algorithm to get maximum rush of adrenaline / money? What will be the space for ethics in future? How today and tomorrow’s societies would be able to avoid a global and general addiction to intense activities such as work, AI, mobile devices, fusion relationships, food, war, … ?
One solution out of this is to engage into jobs and activities that do not need a machine. Playing music with friends, sharing time with kids, making love to a partner, walking in the forest, reading a novel, managing our own lives. But then who will bring the money? Machines? Or humans, still?
With globalization, the world offers lots of possibilities nowadays. Unfortunately, nobody went through all the countries to let you know what would fit the best for your current mood or state.
For example, high PIB countries are usually strongly oriented towards control, and favorite the system versus the individual.
Low PIB countries are rather oriented towards individuals, and care less about macro-economical aspects. It also means less blame and less requirements for high efficiency.
Therefore, if you live a difficult moment, instead of being heavy on your employer and being criticised by your colleagues, think about going abroad for some weeks / months, before coming back. This will enable you to be less blamed and to recover more quickly, prior to come back to highly performing countries, where tolerance for failure is frankly lower.
All I know is that I don’t know anything. Socrates.
Socrates’ statement keeps our hearts and brains busy every day. Still numerical modeling enables to perform predictions at reduced cost. In our recent paper, we have prepared prediction of harmonic spectrum, prior to performing the experiment, and compared both. The agreement was outstanding and was published in Physical Review Letters as editor’s suggestion.
After moving to a new home due to steep increase of rents in Prague, a good fresh up and a new haircut are needed, along with planning the traveling agenda of 2026.
We plan to attend several conferences this year:
On the wish-list, there is also St Jacques de Compostelle, but for personal purposes.