It's insane right now browsing tiktok and seeing every 3rd or 4th video on my feed have consistent glitches (in the same places of a video, as if my internet was slow which it isn't), on only the videos mentioning resistance topics from the US. Very black mirror. Feels like it's meant more to send a message that your content is flagged, and to watch what you say and do. Otherwise they would just block it or hide it from your feed.

Did the consensus shift about TikTok? I thought it was a given that as tech/IT people TikTok isn’t an app worth having on your phone due to spyware/attention brute forcing/curated propaganda by Chinese government.

There are people with a wide variety of opinions here. TikTok is one of the most popular social media platforms, so naturally a lot of folk here will use it too.

And a thread about something on TikTok will naturally select participation from people who have first hand experience or care about it.

It's noteworthy at this point in time that there is a contradiction. The government is currently ramping up Palantir and they are using "precise targeting" of illegal aliens using "advanced data/algorithms". And yet, at the very same time we are seeing time and time again that ICE/DHS agents are finding the wrong people, seemingly going to any house indescriminently, and generally profiling people instead of using any intelligence whatsoever.

Maybe now is exactly the right time to publicly call out the apparent uselessness of Palantir before they fully deploy their high altitude loitering blimps and drones for pervasive surveillance and tracking protestors to their homes.

(My greater theory is that the slide into authoritarianism is not linear, but rather has a hump in the middle where government speech and actions are necessarily opposite, and that they expect the contradiction to slide. Calling out the contradiction is one of the most important things to do for people to see what is going on.)

I think this is mostly because they don't care about false-negatives. They have forgotten the idea that our justice system was supposed to hold true to: "better a hundred guilty go free than one innocent person suffer" (attributed to Benjamin Franklin).

This can be seen in the case of ChongLy Thao, the American citizen (who was born in Laos). This was the man dragged out into freezing temperatures in his underwear after ICE knocked down his door (without a warrant), because they thought two other men (of Thai origin I think) were living there. The ICE agents attitude was that they must be living there, and ChongLy was hiding them. That being wrong does not cost those ICE agents anything, and that is the source of the problems.

Do you mean false positives? A false negative would be "we checked to see whether Alice was in the country illegally, and the computer said no but the actual answer turned out to be yes".

>think this is mostly because they don't care about false-negatives. They have forgotten the idea that our justice system was supposed to hold true to: "better a hundred guilty go free than one innocent person suffer" (attributed to Benjamin Franklin).

Putting on my pedant's hat here. Franklin may well have said something similar, but the maxim you mention is broadly known as Blackstone's Formulation (or ratio)[0] after William Blackstone[1], another Englishman.

Many sayings are ascribed to Benjamin Franklin. And some of them, he actually said.

[0] https://en.wikipedia.org/wiki/Blackstone%27s_ratio

[1] https://en.wikipedia.org/wiki/William_Blackstone

But they were wrong about the Thai people living there. That's the poster's point. Not that they don't care, but that they were wrong from the get-go because they don't actually have good information.

No, it's pretty clear they don't care and will never care.

They are two points and they are both true.

> And yet, at the very same time we are seeing time and time again that ICE/DHS agents are finding the wrong people, seemingly going to any house indescriminently, and generally profiling people instead of using any intelligence whatsoever.

If the end goal is that the broad, general public are intimidated, then they're not necessarily "finding the wrong people." With the current "semi random" enforcement with many false positives, nobody feels safe, regardless of their legal status. This looks to be the goal: Intimidate everyone.

If they had a 100% true positive rate and a 0% false positive rate, the general population would not feel terrorized.

That's exactly what I'm saying though. I agree that their intent is manufacturing fear and uncertainty.

What I'm saying is that congress and the public should be holding them to their word and asking where all this Palantir money is going if the stated intent of "targeted operations/individuals" is completely misaligned with operational reality.

> we are seeing time and time again that ICE/DHS agents are finding the wrong people, seemingly going to any house indescriminently, and generally profiling people instead of using any intelligence whatsoever.

Generally speaking, that is a tactic of oppression, creating a general sense of fear for everyone. Anyone can be arrested or shot.

Yes obviously, but as my central point was, it is the complete opposite of their narrative of targeted operations backed by data.

ICE/DHS are not NSA, they probably don't share efficiently. All the intelligence services are rivals and duplicate capabilities to some degree.

They are going door to door in the neighborhood I grew up in.

They're bringing in a lot of US citizens here in Minneapolis/St Paul, including a bunch of Native folks.

The sex offender they'd been looking for at ChongLy Thao's house had already been in jail for a year.

The Dept of Corrections is annoyed enough about the slander of their work that they now have a whole page with stats and details about their transfers to ICE, including some video of them transferring criminals into ICE custody https://mn.gov/doc/about/news/combatting-dhs-misinformation/

I am pretty nervous about the possibilities for trampling peoples' Constitutional rights in ever more sophisticated ways, but the current iteration can't even merge a database and then get accurate names & addresses out to field agents. (That doesn't stop the kidnappings, it just makes it a big waste of money as adult US citizens with no criminal record do by & large get released.)

The evidence goes strongly against your claims.

[Citation needed.]

The tech industry is falling in line just as expected. Disgraceful. Don't take orders from a LtCol with zero military experience.

To me it reads the other way around, the big money folks in the tech sector are pushing their influence into the military.

Yeah, this reads as the oligarchy further consolidating power.

You seem to be confused. When the tech industry starts appointing people to the military, it's America that is falling in line to them.

Direct commission is a long standing practice, especially for technical fields like medical and now electronic warfare. Surgeons may direct commission to varying field-grade ranks as well, with bonus structure to be competitive with private practice. Military outsources these technical degrees to bring in blood in these voids.

Electronic warfare is typically done by employed soldiers.

Nobody is taking orders from these guys. They're advisors.

They're advisors with the rank of Lieutenant Colonel.

Sure. The US military has done this for decades to prevent brain-drain around emerging technologies to the private sector.

They could make them plain old Lieutenants for that, yes?

Distinction without a difference. They're advisors.

Then why do they need the elevated rank?

I have a dumb question but is it possible to arrange exec-tier or at last officer-tier pay for something like a private?

No level of normal military pay is gonna be meaningful to these folks, at any rank.

They're right above Privates, so still no one (of consequence) is taking orders from them.

What? Per https://www.army.mil/ranks/, they outrank all enlisted and warrant officers, plus Second Lieutenants, First Lieutenants, Captains, and Majors.

Only full Colonels and the 1-5 star Generals outrank them.

You assume they don't agree.

I've been on a similar path. Will have 1000 skills by the end of this week arranged in an evolving DAG. I'm loving the bottoms-up emergence of composable use cases. It's really getting me to rethink computing in general.

Interesting. Could you provide a bit more detail on how the DAG emerges?

2026 paper titled Evolving Programmatic Skill Networks, operationalized in Claude Code

how are they stored?

To tech leaders and hiring managers at other companies: If you're reading this, please consider publicly stating that your company will interview Palantir engineers who want to exit on moral grounds. Create an explicit off-ramp. Lower the barrier to leaving. Make it a tech industry norm that we offer refuge to engineers trying to do the right thing.

Why shouldn't I do quite the opposite? I don't want people with a questionable morale who knowingly built those systems work in my company

The options are a) they have to decide between starving their family or continuing compromise their morals and increasing the capabilities of immoral company X, or b) a more ethically aligned company removes them from the resource pool of immoral company X. Which world do you prefer?

If they're good enough to be hired to palantir as an engineer, I very much doubt at any point they were desperate.

You're missing my point. If all 'ethical' companies treat all ex-employees of 'unethical companies' as unemployable, they are effectively only going to work at 'unethical companies' regardless of whatever mindset has shifted over time.

I agree they shouldn't become outcasts, but it feels disingenuous to say that the reason for wanting to work for palantir is anything but "they wanted A LOT of money".

It's kind of beside the point. You could argue the change of mind/heart is unlikely, but if indeed they had changed, it would be better to encourage that. Perhaps they were lured by the money, or perhaps some jingoistic impulse, but then the reality of what the company was doing became clearer? Or their world view evolved?

That said, if some ex-Palantir worker was somehow working for UNICEF – to take an extreme example – it would be a little awkward unless they had denounced their old company in a fairly public manner.

You could focus on having positive projects for the society, and a good reputation. That works.

I don’t think I ever seen a CV from an ex Pal*ntir employee though. Perhaps they are automatically filtered or working for good morals doesn’t attract them.

I think they might be a little desperate for new employees since I haven’t worked in about ten years and both Palantir and Anduril contacted me with cold calls in past year.

In a country with many huge companies selling oil, cigarettes, weapons, etc. there is no shortage of people willing to deal in morally questionable trades for money. I might even boldly suggest that Palantir is arguably far from the worst.

I can't speak to Palantir, but Anduril is growing rapidly. Headcount has been ~doubling every year.

Palantir had a shit reputation 12 years ago when I graduated from college. I'm not sure folks who couldn't figure that out until now are very principled.

So did HFT's, weapons makers, and Hedge Funds

Select star from blog posts where... :)

Different industry, but our marketing guy once said "You know what this [perfect] metric means? We can never use it in marketing because it's not believable"

Just include some noise, it’s like the most available resource in the universe

Never thought of noise as a resource, but yea.

Thank you for posting binsmith, I've built something similar over the past few days and you've made some great decisions in here

I found a bunch of potential vulnerabilities in the example Skills .py files provided by Anthropic. I don't believe the CVSS/Severity scores though:

| Skill | Title | CVSS | Severity |

| webapp-testing | Command Injection via `shell=True` | 9.8 | *Critical* |

| mcp-builder | Command Injection in Stdio Transport | 8.8 | *High* |

| slack-gif-creator | Path Traversal in Font Loading | 7.5 | *High* |

| xlsx | Excel Formula Injection | 6.1 | Medium |

| docx/pptx | ZIP Path Traversal | 5.3 | Medium |

| pdf | Lack of Input Validation | 3.7 | Low |

Interesting. Was this inspired by the "Context Graphs" concept discussed on X?

No, I don’t hang out at the nazi bar.