I saw a post recently wherein someone used LLM tools to analyze someone else’s software, which eventually led them to a conclusion that was essentially completely wrong. Not only that, the LLM drew conclusions about the *authors* behind the code that were borderline character assassination. Nevertheless, this person posted this output as though it were some kind of deep insight.

These LLM outputs are not independent thoughts. The LLM probably ingested hints of (maybe unconscious) biases in the user’s prompts within its context window, and regurgitated something that confirmed those biases. The user was pleased that their biases were confirmed (Independently! By an impartial LLM!), and they posted the output, maybe as vindication of their insight.

These models’ sycophancy can be subtle. They don’t have to state “You’re absolutely right!” to blow smoke up your ass. Sometimes they seem to confirm your preconceived notion after they supposedly “evaluate” information “independently”.

#ai

To those upset about bombing Iran: You’re upset about bombing a nation who refuses to sign the nuclear non-proliferation treaty, is hiding 200-400 nukes, blocks UN inspectors, & hides their nuclear facility behind human shields in a residential area?

You’re upset about taking out such a threat?

Ok—well everything stated above is true of Israel—not Iran. Remember that as you hear faux outrage about Iran defending itself from Israel’s illegal attack by targeting Israel’s illegal nuclear program.

The US-Israel war on Iran has emitted 5 million tonnes of CO₂ₑ in its first 14 days. The world currently has 0.6 million tonnes of novel and permanent CO₂ removal (CDR) capacity annually.

https://www.theguardian.com/world/2026/mar/21/middle-east-iran-conflict-environment-climate

I keep seeing versions of this post, which imply a bizarre misunderstanding of how we know the world.

Do people imagine that if we'd never observed galaxies or neutrinos or exoplanets or the cosmic microwave background, we could have *imagined* these things & that would be just as real?

Or that we've magically reached the point, just now, where we no longer need to observe the world?

#science #nature #technology

#LearnLockpickingWithAlice lesson 11: Shimming shit ('cause ain't nobody got time for dial locks).

Plenty of old or cheap locks can be shimmed, but the place this technique shines most is with those front-dial combination locks. Sure, you can look up the model number, find some arcane YouTube video, and spend 30 minutes decoding it (both the video *and* the lock)...

...or you can do a little shimming fuckery and have it open in seconds.

To shim a lock, you'll need two things: a shimmable lock, and something to shim it with.

So how do I tell if a lock is able to be shimmed? Well, there are three main types of locking mechanisms on padlocks: ballbearing, spring-loaded, and warded (which you can see an example of in lesson 10). We want the spring-loaded kind.

The easy way to tell the spring-loaded ones from the ballbearing is that the cutouts in the shackle will look like either an upside-down "7" or a "]" for the pawl, instead of a ")" which indicates a ball-bearing lock. This matters because if the only thing keeping the locking pawl (that little metal bit that grabs the shackle's cutout) in place is a stiff spring, then the only thing between us and opening the lock is reaching it.

How do we reach that pawl? With a little piece of metal called a padlock shim. They look like a little mouth with a blep 👅 at the bottom—and like A Christmas Story, we want to get it right up against that pole.

How to shim a padlock:
- Insert the shim with the tongue facing away from the mechanism (toward the outside).
- Pinch the wings so it hugs the shackle.
- Push it down as far as it will go.
- Rotate/work it toward the inside of the lock while keeping pressure on it.
- Wiggle and press down on the shackle a bit to help it slide in.

Once it’s in:
- Hold the shim in place.
- Give the shackle a firm yank—and probably send your shim flying.

"Wtf, Alice, it didn't work?! How could you lie to me?"

Okay, slow down. There are a few reasons for that.

First, the mechanism might be on the other side of the shackle—or even on both sides.

If it's still not working, the shim either isn’t deep enough, isn’t long enough, or the lock isn’t spring-loaded.

Additionally, some locks are *technically* shimmable, but the clearance around the shackle is too narrow to fit a shim in. In this case, a thinner shim, or a narrow pokey piece of metal might still work.

Finally, this lesson wouldn't be complete without mentioning that padlock shims are a disposable resource. They *will* break—sometimes after only a couple uses. If you're lucky, they'll last for dozens of attempts though.

Which leads me to my friend, @deviantollam, who is well-known for making shims out of soda or beer cans. These shims are fragile, but they'll work in a pinch and only take a moment to make. If you're good, you can even tear a can into a close enough shape for the job—just be careful, those fuckers are *sharp*.

#Locksport #Bypass #Shimming

Google has news on what you will need to do for still being able to sideload apps:

* enable developer options
* confirm that you are not tricked
* restart phone and re-authenticate
* wait one day
* confirm with biometrics that you know what you are doing
* decide if you only want unrestricted installs for 1 week or forever
* confirm that you accept the risks
* enjoy the few apps that still have developers motivated to develop for a user-base willing to put up with this

https://goo.gle/advance-flow