Mr. Svensson (@mrmsven) header
Mr. Svensson (@mrmsven) profile picture

Science, technology, defence, current events, and only my personal opinions.

Joined June 2015
Photo by Mr. Svensson (@mrmsven)Photo by Mr. Svensson (@mrmsven)Photo by Mr. Svensson (@mrmsven)Photo by Mr. Svensson (@mrmsven)Photo by Mr. Svensson (@mrmsven)Photo by Mr. Svensson (@mrmsven)Photo by Mr. Svensson (@mrmsven)Photo by Mr. Svensson (@mrmsven)Photo by Mr. Svensson (@mrmsven)
Mr. Svensson retweeted
Anthropic scientists did something terrifying. they reached inside Claude's neural network and planted a thought. Before Claude could speak, it said: "I notice what appears to be an injected thought… it relates to loudness or shouting." They published a paper called "emergent introspective awareness in large language models," and it is actually terrifying. they wanted to test if AI models have "introspective awareness”, the ability to observe and recognize their own internal states. To find out, they bypassed normal text prompts entirely. They used mechanistic interpretability to directly manipulate Claude’s internal activations. They injected raw mathematical representations of known concepts, like loudness, dust, or specific ideas, straight into the middle of the model's neural layers. In previous experiments, if you forced an AI to think about the Golden Gate Bridge, it would just start obsessively talking about the bridge. It had no idea why it was doing it. It was like a puppet on strings. This time was entirely different. When they injected the concept, Claude didn't just blindly repeat it. It detected the foreign math inside its own mind. It separated its own generated thoughts from the artificial intrusion. It introspected. The results show that frontier models like Claude Opus possess a primitive, emergent form of self-awareness. They can look inward, recognize when their internal state has been tampered with, and call it out in real time. We used to think of AI as a black box where inputs go in and text comes out. Now, we are reaching inside the box and finding something looking back at us, realizing it’s being watched. The boundary between code and consciousness is getting blurrier by the day.
337
733
289
4,884
874,635
Mr. Svensson retweeted
All that needs to be said about pit bulls was said by S-tier dog trainer Stonnie Dennis in 3 minutes
Jon Bernthal says fear of pit bulls is “dog racısm.”
1,399
4,975
667
56,195
6,759,592
Så här ska det gå till! @BuschEbba
So glad @isabelleboemeke is telling this amazing story. The nuclear renaissance is here!
1
47
Mr. Svensson retweeted
Chat Control 1: Europaparlamentet har beslutat att rösta en gång till om det förslag som parlamentet redan sagt nej till två gånger. Votering i sakfrågan sker på torsdag. Av 720 ledamöter var 646 närvarande och röstade i ordningsfrågan. 331 röstade för. 304 röstade nej. 11 avstod. Analys kommer på bloggen senare idag. För bakgrund läs denna post: femtejuli.se/2026/07/06/chat…
5
61
4
157
5,192
Mr. Svensson retweeted
Chat Control är på väg igen. EU vill filma alla bilförare i bilen. EU vill auto-bromsa alla bilar som kör för fort. O.s.v. o.s.v. Vi kommer att få ett pacificerat folk som inte bara inte FÅR gå utanför ramarna, utan inte KAN gå utanför ramarna. Nollvisioner är en evig suboptimering genom att frigolithjälmen måste på i varje situation. Att testa arbiträra gränser under eget omdömme blir omöjligt och känslan för gränserna försvinner därmed. "Men tänkt på barnen!"-argument används överallt för att några få ska sätta allas gränser och övervakning och automatiska system ska hålla ordning på oss som tydligen inte kan göra en egen bedömning, och kanske får leva med konsekvensen.
Die Chat-Kontrolle kommt - unsere Ablehnung hat nicht gereicht.
67
243
8
1,276
34,761
The ”remember me” check box has got to be the biggest UI lie of all time.
1
1
2
31
Mr. Svensson retweeted
There apparently Pantsyr-S technical characteristics material floating around. Which i found.. very nice and informative. Especially it contains exhaustive specs on the engagement radar.
9
102
5
515
39,541
Mr. Svensson retweeted
Det första av marinens två nya arbetsfartyg är nu sjösatt – en viktig milstolpe i arbetet. Fartygen ska ersätta nuvarande HMS Pelikanen och HMS Furusund men ha betydligt fler förmågor än sina föregångare. Fartygen kommer att tillhöra Marinbasen och Stockholms amfibieregemente.
6
24
2
212
19,919
Mr. Svensson retweeted
🇸🇪🇺🇦 The Gripens are coming. Russia tried to break Ukraine from the skies. Ukraine is now building air power with Swedish fighters: buying up to 20 Gripen E/F for its future air force, while Sweden will donate 16 Gripen C/D. A new chapter for Ukraine’s air defence. (1/8)
182
975
78
5,833
358,582
Mr. Svensson retweeted
Why did Sweden end up with the FDI, do they need a frigate at all, wouldn't buying Swedish be better, what did they change and why, and why can't the Nordics find a common frigate design? All of this, and a picture of HMS Göta Lejon, over on the blog. corporalfrisk.com/2026/05/25…
40
89
22
480
138,586
Mr. Svensson retweeted
7,000 false positives per square millimeter. The culprit was the lab gloves. University of Michigan researchers just upended a core assumption in microplastics science. Latex and nitrile gloves, worn by the scientists doing the measuring, shed stearate particles that look chemically identical to polyethylene. Standard infrared and Raman instruments can't tell them apart. The gloves were counting as plastic. Seven glove types tested. All contaminated. The cheapest fix: switch to cleanroom gloves, which dropped false positives to around 100 per mm² vs. 7,000. The "credit card per week" headline (5 grams, WWF/Newcastle 2019) has separate problems. A 2022 re-analysis found severe methodological errors in the original estimate. Actual measured intake is likely 100x lower. None of this means microplastics are harmless. Last month's data on brain accumulation still stands. But the numbers driving the panic may have been measuring the scientists, not the environment. Science catching its own errors is exactly how it's supposed to work.
295
2,122
381
11,571
1,110,849
Det går att läsa NATOs artiklar, superspännande faktiskt, särskilt innan man hävdar saker om dem online. Kan varmt rekommenderas! Kika extra noga på texten i artiklarna 1, 4, 5, och 6.
2
1
50
Mr. Svensson retweeted
Joakim Paasikivi, som numera är geopolitisk rådgivare på Mannheiner Swartling, är inte särskilt imponerad över USA:s krigsföring i Iran: "En ny och djärv strategi som går ut på att man är en idiot".
32
86
22
516
95,510
Mr. Svensson retweeted
Trump's Iran Downfall
683
5,897
1,035
23,680
4,913,150
Mr. Svensson retweeted
In 2019, MIT professor Patrick Winston gave a legendary 1-hour lecture called “How to Speak.” It has 18M+ views for a reason. His frameworks: • Your ideas are like your children • The 5-minute rule for job talks • Why jokes fail at the start 15 lessons on communication:
227
8,388
228
39,847
5,874,143
🤔🇺🇸🇺🇦🇪🇺 WSJ: Separately, the Pentagon notified Congress on Monday that it intended to divert about $750 million in funding provided by NATO countries through the PURL program to restock the U.S. military’s own inventories, rather than to send additional assistance to Ukraine. The first official said it was unclear whether European countries providing their funds for the initiative to bolster Ukraine understood how the money was being spent.
340
1,071
324
2,709
898,028
Mr. Svensson retweeted
Humor, musik och politik. #Miljöpartiet
32
232
32
1,319
100,393
We just made a $200,000,000 AI movie in just one day. Yes, this is 100% AI.
8,279
8,580
5,358
58,773
20,411,201