Digital Futures is a cross-disciplinary research center with the vision to solve societal challenges through digital transformation. It was established in 2020 by KTH, Stockholm University and RISE, based on significant long-term support of a Strategic Research Area by the Swedish Government. The vision of Digital Futures is to shape an economically, environmental and socially sustainable society through digital transformation. Our mission is to create an international breeding ground to excel in research, empower people by fostering innovation, education, leadership and entrepreneurial spirit and build a strong community.
On this channel, we share recorded seminars hosted by the center, material from workshops, and news from our cross-disciplinary research. You can find upcoming events on our website www.digitalfutures.kth.se/events/ and you are most welcome to contact us to suggest future seminars.
Isn't infrared in the range of 300 GHz to 400 THz? I have worked a fair bit with fibre optics and recall the wavelengths used are in the vicinity of 1000 nM or 1 µM, with visible light starting around 400 THz. 1 µM corresponds to a frequency of 300 THz.
The button can be called, "I need an adult." 😊 In all seriousness I would explain exactly what the Internet is to my child before they can use it. I would tell them what ego based ad platforms are before they can use Facebook or whatever else and it's not that hard to explain it to your 13 year old. "Hey honey, you know Mommy likes to do scrap booking? Facebook is like that but a stranger sells the information and uses it to try to make you buy things and assesses your value as a human in general. They compare your pictures and messages to everyone else's on their platform and sell the information. They also have security features where sometimes they may even conclude you might be trying to kill yourself so they call the authorities for your safety." "How do they know what I might do in the future?" "They don't honey, they make guesses and stick by their projections about all sorts of metrics. They don't actually care when they are wrong." "Oh... but they have privacy buttons." "Those are just for other users and people on the Internet. Not objective concepts." "Okay. I don't want to sell my pictures." "Great, and you don't have to use these services no matter what people tell you honey. Always remember that." Not that difficult.
Merci matthieu, un peu compliqué de passer outre ton magnifique accent les 5 premières minutes mais une fois qu'on est dedans c'est vraiment agréable, tu abordes un sujet relativement technique pour moi mais tu as réussi à ne pas me perdre bravo ! Si je peux me permettre une petite question, comment peux tu être sur que les paramètres que tu choisis de mesurer sont les plus représentatifs de ton système, et si seulement ils le sont ? Bonne continuation et encore merci
ahah, desole... Mon accent est un peu meilleur aujourd'hui mais c'est quand meme pas terrible. Ta question est tres complexe. Il y a de nombreux travaux sur ce que l'on doit mesurer pour avoir une unique solution (autrement dit, si on ne mesure pas ce qui faut, on risque de ne pas retrouver la bonne solution !). Dans notre cas (le traffic routier), le probleme a ete etudie par Maria Laura Delle Monache (Traffic reconstruction using autonomous vehicles) et elle montre que les mesures proposees ne menent qu'a une unique solution (au bout d'un certain temps). J'espere avoir repondu a ta question !
Parkinson disease is a very terrible illness🦠, my Dad suffered from it for 19 years until we finally got a help and a medicine that truly works that helped treat, cure and reversed all his symptoms... My Dad is completely okay and healthy now💯!!!
Dear DIGITAL FUTURES and KTH team members. Thank you for your interest in INDUSTRY 5.0. My name is Michael Rada, I am the Founder of iNDUSTRY 5.0 implementing its principles in companies and businesses since 2013, leading a global network in 117 countries. If any questions about INDUSTRY 5.0 feel free to ask.
Thanks a lot, this was a great talk (besides the audio problems). Your slide about the derivation of score over time (ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-f0FwTGk3MUk.html) is very interesting. Humans are linear in your graph, too, right? Would humans be able to learn when you deprive them of new data? I don't think so, s getting new data over time must be allowed, right? What exactly are you measuring by score along the time axis? Is an intelligent system by your definition one that increases its score (some model performance metric) along time for one specific problem, getting new data over time? (Machine) learning's derivation of score over time is not >0 in the limit because you set compute to be constant, so there is a limit of data for retraining/fine-tuning/few-shot training? But shouldn't we look at compute not in the sense of complexity classes but FLOPS? Isn't then ML derivation of score over time >0 in the limit? Shouldn't we also look at multiple problems and give intelligence credits to a system if it can generalize to newer and newer problems? If you consider newer and newer problems that some LLM can generalize to (with the help of zero or a few shots), isn't its derivation of score over time >0 in the limit? Finally, why is "super-intelligence" not possible?
Yes, there are several Digital Futures fellows, postdoc fellows, and projects on this topic. You can find a list of people here: www.digitalfutures.kth.se/about/all/ And different types of projects here: www.digitalfutures.kth.se/research/
Hello everyone I want to appreciate Dr. Ekpen for helping me with his herbal medicine that cure my HSV completely. I have been living with HSV for the past 4years. I was so happy when I found Dr Ekpen blogs. the great herbalist doctor who cure my HSV recently. today I'm happy because am free from HSV finally. I was recommended to Dr Ekpen by someone that was cure from HIV last January with the natural herbal medicine of Ekpen, Dr. Ekpen have the gift to STS,leaukemia, Tuberculosis,psoriasis, cure Diabetes, ,herpes virus, code sores .Acne scar, Eczema,pile, tinitus , fibroids, epilepsy, chronic disease, Genital warts, fruit of womb,High blood pressure,for more treatment information you can search him on his RU-vid video, Dr Ekpen. Or Add him up on his Facebook page; facebook.com/Dr-Ekpen-Herbal-Medicine-102532145172128/
You know that you can swap what is adjusted in a neural net. You can use fixed weights and parametric (adjustable) activation functions. The fixed systems of weighted sums can be done using fast transforms like the fast (Walsh) Hadamard transform. The resulting neural nets are very fast and behave very well as autoencoders.-- Fast Transform fixed-filter-bank neural networks -- AI462 neural networks.