Live data from Hacker News

Local LLMs versus offline Wikipedia

evanhahn.com

151–160 of 200 posts

Re: Local LLMs versus offline Wikipedia

#151
post #75

Earlier quoted context omitted.

Remember the first time you touched a computer, the first game you ever played or the first little script you wrote that did something useful. I imagine this is how a lot of people feel when using LLM's especially now that it's new. It is the most incredible technology ever created by this point in our history imo and the cynicism on HN is astounding to me.

> It is the most incredible technology ever created by this point in our history imo and the cynicism on HN is astounding to me. TBH, I still think LLMs have a long way to go to catch up to the technology of wikipedia, let alone the internet. LLMs at their peak are roughly a crappy form of an encyclopedia. I think the interactivity really warps peoples perspective to view it as more impressive, but it's difficult to…

But practically speaking they're probably way more valuable in the start from scratch scenario.

Wikipedia articles sometimes have a lot of jargon, making the information useless unless you have a prior understanding of the subject matter.

Re: Local LLMs versus offline Wikipedia

#152
post #4

This is a sensible comparison. My "help reboot society with the help of my little USB stick" thing was a throwaway remark to the journalist at a random point in the interview, I didn't anticipate them using it in the article! https://www.technologyreview.com/2025/07/17/1120391/how-to-r... A bunch of people have pointed out that downloading Wikipedia itself onto a USB stick is sensible, and I agree with them. Wikipedi…

I've been carrying around a local wikipedia dump on my phone or pda for quite a bit more than 10 years now (including with pictures for the last 5 years). Before kiwix and zim, I used tomeraider and aard. I do it both for disaster preparedness but also off-line preparedness. Happens more often than you'd think. But I have been thinking about how useful some of the models are these days, and the obvious next step to m…

How do you maintain updates? One thing which of concern is rogue edits getting downloaded, have you figured out a mitigation?

Re: Local LLMs versus offline Wikipedia

#153
post #147

Earlier quoted context omitted.

I have been involved with AI for over 40 years. I assure you anyone being shown a current frontier model in operation 10 years ago would have been blown off their socks. Yet here we are. Rather than exploring this fantastic new tool, so many here are obsessed with pointing out flaws and shortcomings. I get the angst of a world facing dramatic change. I don't get the denial and deliberate ignorance flaunted as somehow…

> Yet here we are. Rather than exploring this fantastic new tool, so many here are obsessed with pointing out flaws and shortcomings. Now think about any technology you disapprove of, and imagine that defence: “We have just invented bombs and killer drones, yet rather than exploring these fantastic new tools, so many here are obsessed with pointing out flaws and shortcomings.” > I get the angst of a world facing dram…

I don't think I explained it well if that is what you get from it.

When I say 'I get the angst', I do not mean ungrounded fears. e.g. Captured regulation killing off open model creation and use and locking AI behind a few aligned actors making sure the tech's advantages go to the select few and their serves being one of them. When I say 'dramatic change' I do not mean dramatic as in a comedy play, but real deep societal impact with a significant chance of total turmoil.

What I tried to address is the dismissive 'reactionary' response of belittling and denying the technology itself, not just in some 'tech' circles, but almost endemic in academia. "It's nothing new", "just a 'stochastic parrot'", "just lossy compression", "just a parlor trick", "a useless hallucination merry-go-round", "another round of anthropomorphism for the gullible" etc. etc.

Re: Local LLMs versus offline Wikipedia

#154
post #147

Earlier quoted context omitted.

> Yet here we are. Rather than exploring this fantastic new tool, so many here are obsessed with pointing out flaws and shortcomings. Now think about any technology you disapprove of, and imagine that defence: “We have just invented bombs and killer drones, yet rather than exploring these fantastic new tools, so many here are obsessed with pointing out flaws and shortcomings.” > I get the angst of a world facing dram…

I don't think I explained it well if that is what you get from it. When I say 'I get the angst', I do not mean ungrounded fears. e.g. Captured regulation killing off open model creation and use and locking AI behind a few aligned actors making sure the tech's advantages go to the select few and their serves being one of them. When I say 'dramatic change' I do not mean dramatic as in a comedy play, but real deep socie…

Thank you for the clarification. That did help to understand your specific complaints better.

Re: Local LLMs versus offline Wikipedia

#155
post #143
post #7

One important distinction is that the strength of LLMs isn't just in storing or retrieving knowledge like Wikipedia, it’s in comprehension. LLMs will return faulty or imprecise information at times, but what they can do is understand vague or poorly formed questions and help guide a user toward an answer. They can explain complex ideas in simpler terms, adapt responses based on the user's level of understanding, and…

Sounds like a good way to ensure society never “reboots”. A “frozen snapshot” of reliable knowledge is infinitely more valuable than a system which gives you wrong instructions and you have no idea what action will work or kill you. Anyone can “explain complex ideas in simple terms” if you don’t have to care about being correct. What kind of scenario is this, even? We had such a calamity that we need to “reboot” soci…

Currently, there are billions of devices that are capable of storing and running a 4B LLM locally. Hundreds of millions for 32B LLMs. It would take an awful lot of effort to destroy all of that.

If you're doomsday prepping, there's no reason not to have both. They're complimentary. Wikipedia is more reliable, but also much more narrow in its knowledge, and can't talk back. Just the "point someone who doesn't know what he's dealing with in a somewhat sensible direction" is an absolute killer feature that LLMs happen to have.

Re: Local LLMs versus offline Wikipedia

#156
post #7

One important distinction is that the strength of LLMs isn't just in storing or retrieving knowledge like Wikipedia, it’s in comprehension. LLMs will return faulty or imprecise information at times, but what they can do is understand vague or poorly formed questions and help guide a user toward an answer. They can explain complex ideas in simpler terms, adapt responses based on the user's level of understanding, and…

As a bonus the LLM can spew out endless amounts of bullshit.

Re: Local LLMs versus offline Wikipedia

#157
post #4

This is a sensible comparison. My "help reboot society with the help of my little USB stick" thing was a throwaway remark to the journalist at a random point in the interview, I didn't anticipate them using it in the article! https://www.technologyreview.com/2025/07/17/1120391/how-to-r... A bunch of people have pointed out that downloading Wikipedia itself onto a USB stick is sensible, and I agree with them. Wikipedi…

Someone should start a company selling USB sticks pre-loaded with lots of prepper knowledge of this type. In addition to making money, your USB sticks could make a real difference in the event of a global catastrophe. You could sell the USB stick in a little box which protects it from electromagnetic interference in the event of a solar flare or EMP. I suppose the most important knowledge to preserve is knowledge abo…

The US government already does this. Presumably, many governments do, but I've only ever worked for the US, so it's the only one I know of. Every day, the NSA does a dump of Wikipedia, the Stack Exchange network, and God knows what else to import into self-hosted versions of clone sites on classified networks, so US intelligence and military personnel can access this information without needing an Internet connection. The places these get hosted are already inside of military installations, in SCIFs that are behind several-foot thick concrete and radiation shielding that is probably quite a bit more likely than you to survive some kind of event that otherwise collapses civilization. They, of course, also have all of the military field manuals and technical manuals that more or less form a complete guide to how to survive in the wild with no equipment.

That said, I still think I understand why individuals like to do this kind of thing. You're not really concerned about human civilization itself preserving its structures and knowledge. You're concerned about the possibility that you personally will survive some civilization ending event and whatever is left of global militaries and various larger-scale data archiving systems won't care about you or have any way to share the information.

Just be warned, as someone with past experience being in the military and having to actually do these "remote survival with no gear" things, just reading about it is typically not enough to succeed on your first try. You need practice, and it helps quite a bit to have friends, co-workers, some sort of trusted companions who have at least as much and ideally more experience than you. Whoever figures out how to build the first new piece of "technology X" after catastrophe wipes out the last one we had before is far more likely to be someone who built this kind of thing before than someone who spent the pre-apocalypse data hoarding but never actually practicing what they're trying to learn how to do.

Re: Local LLMs versus offline Wikipedia

#158
post #75

Earlier quoted context omitted.

Remember the first time you touched a computer, the first game you ever played or the first little script you wrote that did something useful. I imagine this is how a lot of people feel when using LLM's especially now that it's new. It is the most incredible technology ever created by this point in our history imo and the cynicism on HN is astounding to me.

I have been involved with AI for over 40 years. I assure you anyone being shown a current frontier model in operation 10 years ago would have been blown off their socks. Yet here we are. Rather than exploring this fantastic new tool, so many here are obsessed with pointing out flaws and shortcomings. I get the angst of a world facing dramatic change. I don't get the denial and deliberate ignorance flaunted as somehow…

This thread is not about flaws and shortcomings. I use Claude code all the time, it's great, it's fun. But "the most incredible technology ever created by this point in our history" (OP quote, we assume "our history" means "human history", as opposed to "history of the past couple of years in the Valley-scape, sure), please. This is a delusional and dangerous point of view.

Re: Local LLMs versus offline Wikipedia

#159
post #123
post #7

One important distinction is that the strength of LLMs isn't just in storing or retrieving knowledge like Wikipedia, it’s in comprehension. LLMs will return faulty or imprecise information at times, but what they can do is understand vague or poorly formed questions and help guide a user toward an answer. They can explain complex ideas in simpler terms, adapt responses based on the user's level of understanding, and…

that's assuming working computers or phones are sill around. a hardcopy of wikipedia or a few selected books might be a safer backup. otoh, if we do in fact bring about such a reboot then maybe a full cold boot is what's actually in order ... you know, if it didn't work maybe try something different next time.

That's a very safe assumption. There are more smartphones on Earth than there are humans.

Re: Local LLMs versus offline Wikipedia

#160
post #141
post #9

Earlier quoted context omitted.

An unreliable computer treated as a god by a pre-information-age society sounds like a Star Trek episode.

“Computer, raktajino”, asked the president of the United Earth for the last time. One sip was followed by immediate death. The new versions of replicators and ship computers were based on ancient technology called LLMs. They frequently made mistakes like adding rusty nails and glue to food, or replacing entire mugs of coffee with cyanide. One time they encouraged a whole fleet to go into a supernova. Many more disast…

> Computer, raktajino”, asked the president of the United Earth for the last time. One sip was followed by immediate death.

Obviously, raktajino would already be programmed in and called via a tool call. The president may get an occasional vodka instead, but will live.

Post reply on HN