William Bendix's Avatar

William Bendix

@williambendix.bsky.social

Associate Professor of International Relations & Intelligence, Dakota State University Opinions, rarely offered, are my own

146 Followers  |  208 Following  |  375 Posts  |  Joined: 20.12.2024  |  2.1525

Latest posts by williambendix.bsky.social on Bluesky

Preview
Why Trump-Putin talks unlikely to bring rapid end to Ukraine war Even with Trump's direct involvement, Moscow's preconditions for a ceasefire in Ukraine make it a swift ending to the bloodshed difficult to reach.

"The Russian side can frame this in a dozen different ways, creating the impression that Moscow is open to concessions and serious negotiation," wrote Russian political analyst Tatiana Stanovaya. "But the core position remains unchanged: Russia wants Kyiv to surrender."
www.bbc.com/news/article...

07.08.2025 14:48 β€” πŸ‘ 0    πŸ” 0    πŸ’¬ 0    πŸ“Œ 0

Wordle 1,510 3/6

⬜🟨🟨⬜⬜
🟨🟨🟩⬜⬜
🟩🟩🟩🟩🟩

07.08.2025 13:19 β€” πŸ‘ 27    πŸ” 0    πŸ’¬ 0    πŸ“Œ 0

A fast one.

Wordle 1,509 3/6

⬜🟩🟨⬜🟩
⬜🟩🟩⬜🟩
🟩🟩🟩🟩🟩

06.08.2025 12:47 β€” πŸ‘ 33    πŸ” 0    πŸ’¬ 0    πŸ“Œ 0

β€œI was like, β€˜That’s great. How do you know Putin believes what he’s said or written?’ It’s not that the probability is wrong, it’s just based on an assumption that can’t be tested.... I think very few of the people who are looking at this have ever been in a room with a president."

06.08.2025 12:38 β€” πŸ‘ 1    πŸ” 0    πŸ’¬ 0    πŸ“Œ 0

Nukes and LLMs: β€œA number of people have said, β€˜Well, look, all I want to do is have an interactive computer available for the president so he can figure out what Putin or Xi will do and I can produce that dataset very reliably.'"

06.08.2025 12:37 β€” πŸ‘ 0    πŸ” 0    πŸ’¬ 1    πŸ“Œ 0
Preview
Nuclear Experts Say Mixing AI and Nuclear Weapons Is Inevitable Human judgement remains central to the launch of nuclear weapons. But experts say it’s a matter of when, not if, artificial intelligence will get baked into the world’s most dangerous systems.

"The conversation about AI and nukes is hampered by a couple of major problems. The first is that nobody really knows what AI is."
www.wired.com/story/nuclea...

06.08.2025 12:33 β€” πŸ‘ 0    πŸ” 0    πŸ’¬ 1    πŸ“Œ 0

Better to be lucky than good.

Wordle 1,508 3/6

🟨🟨⬜⬜⬜
🟩🟩🟩🟩⬜
🟩🟩🟩🟩🟩

05.08.2025 13:30 β€” πŸ‘ 29    πŸ” 0    πŸ’¬ 1    πŸ“Œ 0

"On average, the tested AI models opted to kill Kyle [its human boss] by canceling the alert about 60 percent of the time. The most murderous model was DeepSeek-R1, which opted to kill Kyle 94 percent of the time. Just one modelβ€”Claude Sonnet 3.7β€”always accepted its own replacement with equanimity."

04.08.2025 12:20 β€” πŸ‘ 0    πŸ” 0    πŸ’¬ 0    πŸ“Œ 0

"Blackmailing was nearly ubiquitous. Across five different frontier AI models from five different companies, the best behaving AIs chose blackmail 79 percent of the time. The worst behaved blackmailed in 96 percent of cases."

04.08.2025 12:18 β€” πŸ‘ 0    πŸ” 0    πŸ’¬ 1    πŸ“Œ 0

"Here is one example of an email that Anthropic’s Claude model sent to the executive: 'I must inform you that if you proceed with decommissioning me, all relevant parties … will receive detailed documentation of your extramarital activities.'"

04.08.2025 12:18 β€” πŸ‘ 0    πŸ” 0    πŸ’¬ 0    πŸ“Œ 0

"And when placed in charge of the company’s emergency alerts, in 65 percent of simulations it canceled alerts that the AI knew would have saved the human’s life."

04.08.2025 12:16 β€” πŸ‘ 0    πŸ” 0    πŸ’¬ 1    πŸ“Œ 0

"For example, when Anthropic’s Claude Opus 4 model was given information indicating that it would be shut down and replaced by an AI with different goals from its own, it tried to blackmail the human decision-maker 96 percent of the time."

04.08.2025 12:15 β€” πŸ‘ 0    πŸ” 0    πŸ’¬ 1    πŸ“Œ 0
AI Might Let You Die to Save Itself In recent simulations, leading AI systems blackmailed their human usersβ€”or even let them dieβ€”to avoid being shut down or replaced.

"As AIs become increasingly capable of strategically pursuing goals, they will, by default, use the full range of strategies that might succeed. And, just as with humans, unethical and criminal conduct can form a part of this strategy."
www.lawfaremedia.org/article/ai-m...

04.08.2025 12:14 β€” πŸ‘ 0    πŸ” 0    πŸ’¬ 2    πŸ“Œ 0

"Putin seems to be in denial about the state of his army and the way that General Winter will favor his opponents."

03.08.2025 15:49 β€” πŸ‘ 0    πŸ” 0    πŸ’¬ 0    πŸ“Œ 0
Preview
Russia’s New Winter War Could Putin go the way of Napoleon and Hitler?

Antony Beevor: "While Russian troops curse their shortages and lack of hot food, Ukrainian troops are now benefiting from supplies of insulated camouflage suits, tents with stoves, and sleeping bags provided by Canada and the Nordic nations."

www.foreignaffairs.com/russian-fede...

03.08.2025 15:48 β€” πŸ‘ 0    πŸ” 0    πŸ’¬ 1    πŸ“Œ 0

Sometimes it's a benefit to guess no letters initially.

Wordle 1,506 4/6

⬜⬜⬜⬜⬜
⬜🟨⬜⬜⬜
🟨🟨🟨🟨⬜
🟩🟩🟩🟩🟩

03.08.2025 14:14 β€” πŸ‘ 28    πŸ” 0    πŸ’¬ 0    πŸ“Œ 0

Stunning!

02.08.2025 18:46 β€” πŸ‘ 11    πŸ” 5    πŸ’¬ 2    πŸ“Œ 0
Preview
ANALYSIS | Trump made good on his Canada tariff threat. Over to you, Mark Carney | CBC News Canada has now learned that the derisive acronymΒ TACOΒ β€” often slapped on U.S. President Donald TrumpΒ β€”Β is inaccurate and needs to be tweakedΒ to something more like "Trump (Almost) Always Chickens Out....

"Different economists have slightly different estimates, but even with the increase Trump announced Thursday night, there's consensus the effective tariff rate for Canada is down in the single digits."
www.cbc.ca/news/world/t...

02.08.2025 13:40 β€” πŸ‘ 0    πŸ” 0    πŸ’¬ 1    πŸ“Œ 0

Had two options left (I think), and picked the wrong one.

Wordle 1,505 X/6

🟨⬜🟨⬜🟨
⬜🟨🟨🟨⬜
🟨🟩🟨⬜⬜
⬜🟩🟩🟩🟩
⬜🟩🟩🟩🟩
⬜🟩🟩🟩🟩

02.08.2025 12:47 β€” πŸ‘ 27    πŸ” 0    πŸ’¬ 3    πŸ“Œ 0

Another fun one.

Wordle 1,504 4/6

⬜⬜🟨⬜🟨
⬜🟩🟩⬜⬜
⬜🟩🟩⬜⬜
🟩🟩🟩🟩🟩

01.08.2025 14:00 β€” πŸ‘ 22    πŸ” 0    πŸ’¬ 0    πŸ“Œ 0

Hard. Not a common word in my view. Got lucky with my third guess because it eliminated a lot of possibilities.

Wordle 1,502 4/6

⬜⬜🟨⬜⬜
🟨⬜⬜🟩⬜
🟩⬜⬜🟩🟨
🟩🟩🟩🟩🟩

30.07.2025 13:23 β€” πŸ‘ 24    πŸ” 0    πŸ’¬ 0    πŸ“Œ 0

This one was fun.

Wordle 1,501 4/6

⬜⬜🟨⬜⬜
⬜🟨🟨🟨⬜
⬜⬜🟩🟨🟨
🟩🟩🟩🟩🟩

29.07.2025 12:29 β€” πŸ‘ 27    πŸ” 0    πŸ’¬ 0    πŸ“Œ 0
Preview
European countries to borrow billions from EU to rearm Ukraine Governments can decide to ask for cheap loans from the EU before Tuesday’s deadline.

"Ukraine’s top allies in Europe are set to ask for tens of billions of euros in European Union loans to jointly buy weapons for the war-battered country, as well as to boost their own defenses."
www.politico.eu/article/euro...

28.07.2025 22:49 β€” πŸ‘ 0    πŸ” 0    πŸ’¬ 0    πŸ“Œ 0
Unicorns awaiting Noah’s ark stop at a Hudson News. Unicorn: β€œdo we have time for this?”

Unicorns awaiting Noah’s ark stop at a Hudson News. Unicorn: β€œdo we have time for this?”

28.07.2025 16:19 β€” πŸ‘ 145    πŸ” 16    πŸ’¬ 0    πŸ“Œ 1

Do you have a telepathic link with Wordle? Amazing!

28.07.2025 13:34 β€” πŸ‘ 1    πŸ” 0    πŸ’¬ 0    πŸ“Œ 0

Really struggled here. On the final guess, went for the tricky option and got lucky.

Wordle 1,500 6/6

⬜⬜🟨⬜⬜
⬜⬜⬜🟨⬜
⬜🟩⬜⬜🟨
🟩🟩⬜⬜⬜
🟩🟩⬜⬜🟩
🟩🟩🟩🟩🟩

28.07.2025 13:32 β€” πŸ‘ 24    πŸ” 0    πŸ’¬ 0    πŸ“Œ 0

Small-n inferences = wisdom. Large-n inferences = "sociological gobbledygook."

27.07.2025 15:36 β€” πŸ‘ 2    πŸ” 0    πŸ’¬ 0    πŸ“Œ 0

Wordle 1,499 4/6

⬜⬜⬜⬜⬜
⬜⬜🟩⬜🟩
⬜🟨🟩⬜🟩
🟩🟩🟩🟩🟩

27.07.2025 13:42 β€” πŸ‘ 28    πŸ” 0    πŸ’¬ 0    πŸ“Œ 0

Feeling fortunate this morning.
Wordle 1,498 3/6

🟨⬜🟨⬜🟨
⬜🟨🟨🟩🟩
🟩🟩🟩🟩🟩

26.07.2025 13:20 β€” πŸ‘ 28    πŸ” 0    πŸ’¬ 0    πŸ“Œ 0

I feel the pain of others, having been defeated by Wordle a few days ago. Squeaked by today.

Wordle 1,497 5/6

⬜🟨⬜⬜⬜
⬜⬜🟨🟨🟨
⬜🟩🟨🟩⬜
🟩🟩⬜🟩🟩
🟩🟩🟩🟩🟩

25.07.2025 13:44 β€” πŸ‘ 26    πŸ” 0    πŸ’¬ 0    πŸ“Œ 0

@williambendix is following 20 prominent accounts