Reposted
Why AI generated images are going to damage quality defence reporting
AI slop is being used by ‘quality’ and ‘fanboy’ defence blogs and news websites as well as mainstream media. It’s terrible news!
A rant 🧵
AI slop is being used by ‘quality’ and ‘fanboy’ defence blogs and news websites as well as mainstream media. It’s terrible news!
A rant 🧵
January 13, 2026 at 11:05 AM
Why AI generated images are going to damage quality defence reporting
AI slop is being used by ‘quality’ and ‘fanboy’ defence blogs and news websites as well as mainstream media. It’s terrible news!
A rant 🧵
AI slop is being used by ‘quality’ and ‘fanboy’ defence blogs and news websites as well as mainstream media. It’s terrible news!
A rant 🧵
"Grounded in the core tenets of warfighting, intelligence and enterprise operations ... the War Department will accelerate America's Military AI Dominance by becoming an AI-first warfighting force across all domains."
www.war.gov/News/Release...
www.war.gov/News/Release...
www.war.gov
January 14, 2026 at 6:59 AM
"Grounded in the core tenets of warfighting, intelligence and enterprise operations ... the War Department will accelerate America's Military AI Dominance by becoming an AI-first warfighting force across all domains."
www.war.gov/News/Release...
www.war.gov/News/Release...
"Secretary Pete Hegseth announced that Pentagon networks, including classified ones, would enable access to Grok, the Elon Musk-owned, Saudi- and Qatari-backed AI chatbot"
www.defenseone.com/policy/2026/...
www.defenseone.com/policy/2026/...
Grok is in, ethics are out in Pentagon’s new AI-acceleration strategy
Seven projects are to lead the charge to embed artificial intelligence ever more deeply in military affairs.
www.defenseone.com
January 14, 2026 at 6:57 AM
"Secretary Pete Hegseth announced that Pentagon networks, including classified ones, would enable access to Grok, the Elon Musk-owned, Saudi- and Qatari-backed AI chatbot"
www.defenseone.com/policy/2026/...
www.defenseone.com/policy/2026/...
Open-ended wargames with large language models github.com/IQTLabs/snow...
GitHub - IQTLabs/snowglobe: Open-ended wargames with large language models
Open-ended wargames with large language models. Contribute to IQTLabs/snowglobe development by creating an account on GitHub.
github.com
January 4, 2026 at 12:13 PM
Open-ended wargames with large language models github.com/IQTLabs/snow...
Reposted
Paper in the latest issue of the CIA's in-house journal. LLMs in war-gaming. "in testing we observed that responses generated with Mistral 7B frequently leaned toward diplomatic negotiation, even when behavior was inconsistent with the assigned persona." www.cia.gov/resources/cs...
www.cia.gov
January 1, 2026 at 8:38 PM
Paper in the latest issue of the CIA's in-house journal. LLMs in war-gaming. "in testing we observed that responses generated with Mistral 7B frequently leaned toward diplomatic negotiation, even when behavior was inconsistent with the assigned persona." www.cia.gov/resources/cs...
Reposted
I gave myself PTSD collecting 2y of Russia's attacks on us in the Eastern Flank, formatting data FINALLY into a Spreadsheet tinyurl.com/29jwc3n5
• In NOVEMBER, csv was converted by hwt.lv Team to JSON, now a ONE PASTE entry 🥰
• In DECEMBER, ONE CLICK😍
• In NOVEMBER, csv was converted by hwt.lv Team to JSON, now a ONE PASTE entry 🥰
• In DECEMBER, ONE CLICK😍
December 20, 2025 at 2:45 PM
I gave myself PTSD collecting 2y of Russia's attacks on us in the Eastern Flank, formatting data FINALLY into a Spreadsheet tinyurl.com/29jwc3n5
• In NOVEMBER, csv was converted by hwt.lv Team to JSON, now a ONE PASTE entry 🥰
• In DECEMBER, ONE CLICK😍
• In NOVEMBER, csv was converted by hwt.lv Team to JSON, now a ONE PASTE entry 🥰
• In DECEMBER, ONE CLICK😍
Reposted
"The Chinese-made DeepSeek-R1 artificial intelligence (AI) model produces more insecure code when prompts mention subjects considered sensitive to the Chinese Communist Party (CCP), according to recent research from Crowdstrike." www.lawfaremedia.org/article/deep...
DeepSeek and Musk's Grok Both Toe the Party Line
The latest edition of the Seriously Risky Business cybersecurity newsletter, now on Lawfare.
www.lawfaremedia.org
December 2, 2025 at 3:04 PM
"The Chinese-made DeepSeek-R1 artificial intelligence (AI) model produces more insecure code when prompts mention subjects considered sensitive to the Chinese Communist Party (CCP), according to recent research from Crowdstrike." www.lawfaremedia.org/article/deep...