Chemist, blogger and science journalist. Physical sciences editor for Spektrum der Wissenschaft. Writes about chemistry, infectious diseases, earth sciences, disasters and quokkas. Posts in english and german.
29,112 followers 3,154 following 4,696 posts
view profile on Bluesky 
The lovely science community on Bluesky (even without the comparison to the abusive mess on X) has been plain for a long time, but it's nice to have it confirmed by numbers. #science 🧪 #Bluesky #community #socialmedia www.theguardian.com/technology/2...
Many physicists have the amazing skill to actively not understand everything outside physics.
That's what people want to believe: it's _them_ that are responsible and it has nothing to do with _us_
Could it be that the various shortcomings of AI tell us a bit more about humans than most people want to contemplate?
Yes, it's a common issue, it occurs in gravitational wave detection as well. The difference here is that there is a loop with AI generating new real world data and feeding itself with it.
There's an interesting caveat in there: “It generates its own training data and learns from that data to pick precursors that would give the brightest quantum dot” This is vulnerable to a kind of founder effect, where random early results determine the direction of later tests. #chemsky 🧪
Ah, die random Reflexposter von den Propagandabrigaden. 🙄 Wie hab ich sie seit der Pandemie nicht vermisst.
Unfassbar.
Subnautica players support kyanite! ✊️
Boah, jetzt geht das bei uns schon los wie beim HSV! 🤣 Erstma übern Derbysieg freuen und weiter Punkte machen gegen den Abstieg. Bloß nicht übermütig werden. 🫣
Hooray! It's Mineral Cup season! Come for the pretty pictures, stay for the educational smack talk! What bystander fanbase can we pull in this year? #MinCup25
Also keine Katzen? 🥺
😆
Gibts n Größenlimit?
Because people are currently talking about the consequences of a certain technology in isolation. And that's really a problem, because it distorts the real issue. Also, that all issues are ultimately people issues is clearly true, but a bit too general to be helpful, I think.
It really depends on the schale and timeframe of a mass extinction. There are many examples of biological crises in the history of our planet, e.g. the PETM. And absent a really major disaster life recovers very quickly. And no, we are not "the planet".
That's exactly my point.
🤭 Why not both?
That's rather the point, isn't it? And sorry, totalitarian regimes actually do worse, so dictators won't create that "perfect control" either. And the idea that we have any control over the planet is rather optimistic given that even the fastest processes span several human generations.
This is the context of generative AI. It's the next example of humanity rolling out technology at scale without understanding the consequences. It might harm us directly or by societal and planetary changes. This is not because we lose control of either. It's because we never had control. 7/7
All this talk about dangerous superintelligence, control and the planet obscures this very important insight. Humanity is floundering within a technosphere out of our control on a planet we barely understand. There is a lot of stuff that can really hurt us, and I doubt we know all of it. 6/7
AI doesn't need to "surpass us" or escape our control to cause untold harm. There are already more than enough examples how a mere statistical text generator can mess up people and societies. Because we really don't have much control about technology at all. And that's really the point. 5/7
The reality is that we already have very little control about our technology and its consequences. We have no control over the planet and understand precious little of it. Because of that we have caused very dangerous developments that already make us little more than bystanders. 4/7
The idea behind this is that we humans are a superintelligence controlling the planet by our technology. And now a new super-superintelligence will wrest this control from us and causes runaway destruction on "our" planet while we are relegated to the sidelines. But this is rather misleading. 3/7
The biggest danger usually is described as AI "surpassing us" or mankind "losing control" of ever-powerful AI which will then annihilate or enslave us and "destroy the planet". But there is a problematic subtext here regarding our place in the world and in the technosphere we have built. 2/7
So, let's unpack the whole mess of human hubris surrounding everything #AI and why this is so dangerous. The hubris demonstrated by GenAI companies is of course well documented. But it's also very evident in the criticism, and in a way that is distorting the real issues here. 1/7
Das Vagina Museum macht übrigens ganz, ganz tolle Wissenschaftskommunikation! Der ganze Thread ist lesenswert, zum Schmunzeln, aber auch zum Lernen!
🧪
LOL.
Actually I know a lot more about the subject than you do.
No. It's just bad for humans and the world we are accustomed to. That's not "the planet" or "life". The planet will look quite unperturbed again in likely less than a million years, no matter what we do.
No, it's not really. These are distinct categories. AI may well fuck up human civilisation and/or cause a major biological turnover. But that's not "the planet" by far. People just love to pretend that humans are all that matters, so they call our little mess as threat to "the Planet". It's not.
OMG! /o\
Funktioniert bei uns ziemlich gut. Kannste auf Mastodon nur niemandem erzählen, weil man auf Mastodon nicht drauf klarkommt, dass irgendwo irgendwas funktioniert außer auf Mastodon. 😅
Ich kann das Ding Zusammenfassungen, Titelvorschläge, Teaser etc von Artikeln schreiben lassen. Ich les den Text halt vorher. KI transkribiert meine Interviews. Ich hab das Interview selbst geführt. 🤷♂️ Außerdem machen wir automatisierte Alt-Texte zu Bildern, die können wir direkt abgleichen.
Gelegentlich. Ich hab die letzten Monate ziemlich viel damit rumexperimentiert, aber jetzt mach ich nur noch die zwei, drei Sachen, die sich als sinnvoll erwiesen haben.
Also, ChatGPT is not "bad for the planet". A whole continent covered with a kilometer of basaltic lava is bad for the planet. A ten-mile rock coming in at 20 kilometers per second is bad for the planet. Supercontinents are bad for the planet. Just saying.
Many people hate generative AI, for very good reasons. But when you look into it carefully, what they really hate is people. For very good reasons. (disclaimer: I already hated people before they came up with ChatGPT)
No, I distinctly didn't. 😬
...oder generell irgendwas glaubhaft machen, umMenschen zu manipulieren. Sehe ich auch so.
This!
The issue here is IMO that you both are talking about "mind" as a physical entity of sorts that should exist independent of matter. But mind can be a collective property of the brain that is simply different from the sum of its physical parts.
If we talk about a materialistic/physicalistic view, we can answer the original question very easily. AI can't suffer. Actually no one can suffer at all, because suffering is not a measurable quantity. Problem solved. Unfortunately no one will take that seriously.
The Turing-Test was bullshit from the start, really. Since we don't have a useful definition of concepts like intelligence or, for that matter, mind or consciousness, everybody can claim anything. Especially to sell AI. But we really don't have to buy.
Actually it's very doubtful that a brain "executes a program" or does anything even remotely similar to a computer. The idea that we even understand brains enough to make such a claim is just hubris.
Es geht um einen Artikel darüber, ob KI leiden kann.
I won't take it purrsonal.
Look, it's all hype. There is no mind. it's just a machine running sophisticated mathematics algorithms.
Ich will hier gar nix.
🤯
Error Code: Problem Exists Between Chair And Keyboard
Eventuell sind diese "datenbasierten" Ads echt nur ne Vermarktungsstrategie an die Werbekunden/Datenkunden und die Provider wissen seit 15 Jahren, dass es Quark ist.
Siehe meine andere Antwort. Ich halte dieses datenbasierte "maßschneidern" von Ads/Suchergebnissen/Startseiten für ziemlich ineffektiv. Bestes Beispiel: jedes mal zwei Wochen lang Werbung für das Produkt, das man gerade gekauft hat.
Deutsche Politik: der gleiche Scheiß, nur in provinziell.
Weil das "maßschneidern" halt in der Realität nicht funktioniert. Bestes Beispiel: "maßgeschneiderte" Onlinewerbung, die mir zwei Wochen lang das Produkt anbietet, das ich grad gekauft hab.
I honestly believe that many people like Trump because he's the best entertainment running in the media right now.
LOL. 😄 "Netflix has access to an almost unparalleled set of highly tailored data points about every single one of its customers. [...] But, in a way, isn’t it better to just blankly state that all Libras like Peaky Blinders and be done with it? I am willing to say yes it is."
Kommt ja auch nicht direkt überraschend. ^^
Der Artikel ist über die Ozeane. Siehe Titel. Trends an Land sind sehr viel schwieriger zu messen und für die Gesamt-Biomasse relativ sekundär.
Last Call: Wegen Urlaub wird die Anmeldung für Köln diese Woche dicht gemacht. Wer noch dabei sein will, müsst jetzt zu Potte kommen. 🙂 #whiskytasting
It's probably not much consolation for you, but over here in germany many people will indeed be unfamiliar with those Melbourne suburbs.
At first glance it looks obvious that more species diversity also means more biomass in general. But there is no guarantee at all that this relationship holds over geologogical timespans. So it's good someone had a look. :-) 🧪
You know what’s dumber than asking ChatGPT for life advice? Publishing an article titled, “I asked ChatGPT about … Here’s what it said.”
It's important to note that machine learning has been around for decades and is very useful for many tasks. Conflating Generative AI with those only proves that GenAI companies are realizing that their creation doesn't have enough to offer on its own.
A M7.5 earthquake just struck a remote corner of the Antarctic Plate, below the Drake Passage. What’s going on in this fascinating part of the world? We wrote about it last May, after a M7.4 earthquake. We have made that post free to read, and added a few new maps and thoughts. ⚒️ 🧪
Honestly, I like it. I find it rather creative compared to naming the stuff after your PI.
Ich habe geschrieben, dass das Gerede von "woker Cancel Culture" bloß eine Verschwörungstheorie ist, und dass echte Meinungsunterdrückung viel eher von autoritären Regimes ausgeht. Das hat einige Leute schwer getriggert. Und das ist auch gut so. Aber ein paar Anmerkungen sind noch wichtig:
Nur wenn du Polyole haben willst...
😁 Legendär!
Really fun and interesting talk by @morandilab.bsky.social with a really amazing mechanism and a completely unexpected path toward recyclable polymers. #dreamreactions #chemsky
Yay, ein Talk mit Meerwein-Ponndorf-Verley-Reduktion und Oppenauer-Oxidation! 😀🥳🎉 #chemsky #dreamreactions
I can only recommend attending the #dreamreactions conference. Good talks so far, impressive posters and researchers. And you feel that the organisers put a lot of love and effort into it. Also, lots of bananas. 😄 #chemsky
There should be some kind of award for the most disturbing slide of the conference. #dreamreactions #chemsky
Japp. Passt. Stadthotel Münster.
No Conference for Old Men 😅 #dreamreactions #chemsky
Westfalen.
Japp. 🤣 Gleiches Hotel?
She is not known for hosting loud partys.
😁
Westfalen. ^^
In this hotel room in germany a guest once got horribly drunk and died after running into the wall. They put up a cross in his memory that hangs there to this day.
Keine Ahnung. Der Weltgeist hat einen schrägen Humor? ^^
Love the Story. 😄 I think when you become really good at something you probably make less mistakes than before, but you notice many more of them. Also, you stop rationalising them away.
No comment. 😁
Zum Glück merkt es ja auch niemand sonst. Im Gegensatz zu jedem hinterletzten Tippfehler, den dir 500 Leute hinterhertragen.
One of the curious features of being a science journalist is that all the time people complain about alleged errors or want to split hairs about phrasing - but the one day where I get something really wrong and produce high-grade bullshit, absolutely no one notices. 🤷♂️
Julia Klöckner ist wie Fidel Castro. Castro war auch Politiker. Und er hatte auch eine Affäre mit einer Journalistin. Ich seh‘ da keinen Unterschied, nur dass Castro halt links war.
Reducing emissions of any sort doesn't reverse climate change.
It's not. Climate policy has mostly failed so far, but climate change is firmly established in global consciousness. The current attempts to remove climate as a policy issue will not work, and it will return with force. However, the actual outcome is very much open.
Gern. :-) Wie gesagt, ich weiß es nicht definitiv, aber ich vermute, der Wohnort ist da bedeutsamer.
Das ist ein großartiges Buch!
Eigentlich nur konsequent: wenn man die Redaktion mittelfristig durch nen statistischen Textgenerator ersetzt, kann man auch gleich versuchen, das als Mehrwert zu verkaufen.
Puh, das ist aber auch ne hydrologisch anfällige Ecke. 🙈 Ich drück die Daumen für Regen.
I don't think we matter enough even for that. Humans will live out their few ten or hundred thousand years and then life goes on.
😍
Das ist heute ja genau so, das hat sich kein Stück geändert.
I can imagine! 🤭
Yes, but they still grow in an average garden. Try that with wheat or corn.
I don't think so. Tjey won't have the skills and resources to farm modern high-yield varieties or work steel. It will be potatos and pig iron, if they are lucky.
Also there is a bigger reason not to despair: whatever happens, humanity will not do that much damage. So eco-topia will reassert itself in likely much less than a million years. Even the early triassic apocalypse world lasted probably only 10 million years, and that was far beyond what we can do.