Robert (Bob) Kubinec

Many of the worst Middle Eastern dictatorships emerged from instability:

1. Saddam Hussein
2. Muammar Qaddafi
3. Houari Boumedienne
4. Omar al-Bashir
5. Hafez Al-Assad

Just to name a few... 😅

NYT article on khameini successor

If your version of "extremely progressive" is Mohammad bin Salman then OK I guess 🫣

Clearly trying to appease US/Israel so that the guy can stay alive long enough to take over

What is clear from this recent crisis is that #bluesky is the only reasonably reliable source of info. X is dead for that purpose. It's just Grok and weird memes now.

North Africa has the olive oil, so you're not far off

Yes though Claude may already be on the other side of that convo... 🫣

Meet three scientists who said no to Epstein The warning signs included a web search, a mother’s doubts, and inklings of a “sexist attitude”

This is a nice contrast to all the awful behavior confirmed through the Epstein files. Proud of @jenlucpiquant.bsky.social & @seanmcarroll.bsky.social for picking up on the sexism & charlatanism right away.

all you guys on here being like "claude will change the world" and it just totally bombed trying to write code for a problem set in my glm class. basic stuff.

Guys, this is why you go to school. I guarantee your data science consulting firm is not thinking through what it means that the "machine guarantees the truth."

I think this is the boundary for me: in any particular application, does the machine guarantee the truth? If yes, we should outsource it to a machine (traditional statistical computing, simulations, solvers, for example). If not, we should not.

I don’t think we fully grasp how powerfully our politics is shaped by constant threats of violence

The tricky part is restricting the model to only generate free text when you want it to. That's what makes the Claude Code system nice is that it is highly structured to avoid doing that. The more open-ended the task, probably the more room for hallucination/generated text mistakes.

Please Publish In The APSR

Who does this guy think he is, an elected representative of the American people???

I don't think "AI made this" is an excuse. You sign your name to it, that means you wrote it.

Geez

So maybe another way to put it is that it's not about just cognition but creative cognition. Learning some javascript requires cognition but is not an important/creative task for me. So outsourcing to an LLM is fine as who cares if it's "good" javascript.

Of course with the caveats about security

this is also true.

I think what I'd say is that if you use it for something more cognitive/requires expertise, that incurs a cost because you have to check it in some way. So it's best when it can be checked easily or the automation benefit is significant.

I think some kind of asymmetric warfare. This is essentially what Al-Qaeda & its progeny perfected.

Drones had this feature until recently but we've designed some apparently effective safeguards against them. Of course not all places are as well-protected as U.S. military bases.

And that's not to even mention the concerns about LLM hacking/injected code, data privacy, the inevitability of hallucinations, etc. Very few academics are prepared to deal with serious cybersecurity challenges.

I think they are best used for tasks with a minimal amount of thinking/cognition--whatever the digital equivalent is of taking out the trash. Optimal AI usage involves maximizing cognition by humans by using it for tasks that are rote in nature.

AI tools are very useful for academic research, but at the same time skepticism *is* warranted.

One of the side effects of these tools is that they create over-confidence. The underlying models have no uncertainty and the natural language can lull people into imagining the model "understands" them.

‘Destruction is not the same as political success’: US bombing of Iran shows little evidence of endgame strategy As US bombing operations in Afghanistan, Iraq and Libya have shown, destruction is not the same as political success.

“Destruction is not the same as political success.”

US air power can kill leaders and destroy infrastructure, but it can't build up a new government. An international relations scholar explains lessons from Iraq, Libya, Afghanistan and Kosovo.

buff.ly/694z4Ds

I'm hearing about people evacuating the UAE by land, crossing the border into Saudi Arabia where at least the missile strikes are less frequent.

This war is happening far from U.S. shores and Americans can continue their daily lives without issue.

But with a president like this in charge, with no restraints, it's just a matter of time until war comes to our shores too.

We will host a South Asia APSA pre-conference at Harvard on Sept 2.

If you would like to present a paper, serve as a discussant, or only attend the workshop or happy hour, please submit your info by March 15 (extended deadline):
forms.gle/ZDau8Jm86kJt...

Feel free to circulate widely!

The politics are important and all but also a LOT of people have died

I agree with this perspective. AI is not so much undermining research processes as worsening existing flaws, especially with publication & review.

If this ultimately leads to a collapse & rebirth of our existing systems... I don't think that's necessarily a bad thing for science.

The things people offer as examples of LLM disruption in academia often strike me as insights about the way things already work. E.g. “Now we can mass-produce mediocre papers” or “Now the value of an RA is not that they do the grunt work but that they bring a different perspective.”

Civilian cost of US-Israeli war mounts in Iran | The National Flimsy defences and strikes on homes and businesses raise fears for residents

Air defence systems were “not really working” because most had been damaged or destroyed in the 12-day war with Israel last June, or on the first day of this conflict on Saturday, a resident of #Tehran #Iran told me.

“We don’t even have alarms.”

www.thenationalnews.com/news/mena/20...

description of the reasons why a woman did not give vaccines to her children

Such a sad comment on the state of information in the digital era.

In so many ways, digitalization has made us less, not more, knowledgeable.

www.nytimes.com/2026/03/03/u...