SoylentNews

I do that.
I do that not.
I do do that.
I don't do that.
I do do do that.
I don't do do that.
I do do do do that.
I don't do do do that.
...

Most people on here seem pretty convinced that large language models aren't capable of much useful intelligence.

I do think that language can be a powerful tool for reasoning and thinking about facts but as far as I can see there's something fundamentally different about the way humans use language to these AIs.

Specifically, humans get lived sensory experiences of the concepts that they learn words for. When they learn to talk about roses, there's a good chance they'll get to see and touch and smell a rose and they'll have vivid sensory experiences about that. When they learn what a hill is, they might feel the effort of walking up one. Correct me if I'm wrong, but I don't think most of these current chat bots are exposed to any rich sense of meaning of the words; as I understand it they're just spotting patterns between milions of different sentences that include those words. If they can only "understand" words in terms of other words, then that's a pretty poor imitation of what humans mean when they say they understand a word.

I think the same thing goes for Searle's Chinese Room thought experiment as well. Maybe the system as a whole "understands" Chinese in that it knows everything it needs to about translating it into English, but it doesn't have a rich experience of any of those words, so it's a very dull kind of understanding.

Chat bots are probably already being given access to images and text at the same time, but seeing an image of a rose along with lots of sentences about it is still pretty meaningless if you don't have senses and a life living and moving around plants to know what it is in its proper context, how it relates to you and your story.

What a beautiful thing our little cadre of nerd rebels created here together.

The wider world is pretty much screwed by human short-termism and greed but we bucked that trend here by saying "Fuck No" to the relentless pace of corporate-driven change for the worse.

In a rare moment of optimism I previously wrote about creating your own micro-utopia. Perhaps we were well on our way to doing it here.

I want to thank everyone who's a part of this site for your participation here, hell, for your love of this community.

I sincerely believe what we made here is a uniquely great concept that needs to live on.

I feel sure we'll make another one (maybe even still soylentnews.org) -- at the worst I'll probably just make a subreddit myself if no-one else makes this work -- and I can't stand reddit!

So preserving a cybernetic home for us arguably isn't our biggest concern -- it's a big amount of work but there's something more important -- we're a small community and we need to stick together through this and not fragment or give up.

If at all possible, our new site should keep an old skool, lightweight interface; it needs to be be a primarily textual forum with user submitted stories and yeah I think it should have user moderation if possible. We're mostly curmudgeonly greybeards -- best not to scare those away.

More importantly though, I really think we should only try one (or at most two) new sites at a time. If there's a backlash to what we try, if we keep it small and simple we can just change it or try something else (ideally under the same domain of course). If there are too many new sites this little community will fragment and eventually evaporate.

In this increasingly brutal, isolating world, don't underestimate the value of the connections and kinship we have here. I believe as the years go on this will become more and more important to us. Now let's get on with making it work!

P.S. If anyone finds this weirdo worthy enough of keeping in touch with, write to me. acid, immediately followed by an underscore, soycow, and that address resides at proton.me.

[khallow:] They would gain serious opposition throughout the world by alienating a bunch of developing world countries who need that food.

Here, the poster (me) is saying that they believe that certain unnamed countries “need that food”. I later elaborate that “need” means “Egypt would be a smoking ruin, if it ran low on wheat and it's far from the most unstable in that regard.” Here, “need” means must have or some societies, including Egypt, would fall apart into ruin, if they didn’t have enough food. Even if one disagrees with the claim, it’s a fairly honest use of the word. A good is “needed” when there’s an extremely undesirable outcome, if the need is not met.

Case 2:

[AC:] So there's no problem if the food you're making out of the insects etc is from squashed versions.

If you need to do other stuff (like remove the poop etc) then just squash the head really fast.

Here, “need” has a different meaning: an essential step in some process. AC doesn’t specify what the process could be, but it could be mandated by regulation or even merely that the food tastes better without the poop. But the idea is that if bug poop is to be removed, this approach is a way to do that.

Now, let’s consider a couple of less honest uses of “need”.

Case 3:

[AC:] A mega constellation isn't a necessary step or a necessarily shorter path to a future in space.

Here, the word being abused is “necessary/necessarily”. The complaint in question is that megaconstellations have significant externalities – light pollution and possibly space junk. The implication here is that because a megaconstellation isn’t necessary – there are other unspecified ways to a future in space, then it shouldn’t be done.

There’s two flaws in the argument. First, just because something isn’t needed, doesn’t mean we should be blocking it. Another space example is someone arguing that nobody wants to go to space because the speaker doesn't want to go to space.

Second, when a destination is necessary, then so is a path. For example, suppose a kid needs to go home (it’s getting late) and there are two physical paths to their house. A neighbor turns the kid away from the first path because they can go the other way – with the argument that the kid doesn’t need to go down this path (and presumably irk said neighbor). So then another irkable neighbor at the second path does the same, because the kid doesn’t need to travel down that path either since they could travel down the first path. Now, we’ve gone from two paths home to zero paths home!

This is how the need argument can sabotage not just one endeavor, but all of them. There is no path to space that won’t create a bunch of stuff in orbit and engender the externalities, and where there are so many possible paths to space not a one of them is the unique, necessary path.

Finally, there’s the completely bogus use of “need” that spurred this journal. I’ll quote it in context from the original story here [edit: fixed typo].

Case 4:

Ruiter says he’s continued to talk about data centers because he wants to remind people that “the cloud” they’ve come to rely on isn’t just an ethereal concept—it’s something that has a physical manifestation, here in the farmland of North Holland. He worries that growing demand for data storage from people, and also, increasingly, AI, will just mean more and more hyperscale facilities.

“Of course, we need some data centers,” he says. But he wants us to talk about restructuring the way the internet works so they are not so necessary. “We should be having the philosophical debate of what do we do with all our data? I don’t think we need to store everything online in a central place.”

Basically, Ruiter is a politician mooching off Dutch farmer discontent over harsh EU nitrogen regulations which then boiled over to complaints about data centers (which I gather politically are a vastly safer target) which are competing for the same land as the farmers. And he advocates that we restructure Netherlands society so that data centers “are not so necessary”. All this for a naked self-interest – less competition for Dutch farmland. Note also the process would result in significantly fewer data centers and thus a centralization of all that data contrary to the alleged benefit of the scheme. He threw out an excuse for this, ignoring that the scheme would make the excuse worse not better.

This is the cynical, entitled endpoint of the rhetoric of need: you don’t need this so gimme. No cost to society is too high. Just restructure society so it doesn’t need what I stole from it. I find it interesting how so many people are intent on reenacting those cheesy Ayn Rand novels – not as a ruggedly individual John Galt, but as a sleazy, corrupt Wesley Mouch.

[istartedi:] Ethics is easy. We know there are unethical people, and we know that the people who are charged with reigning them in are also unethical. Money is an easy target, but those targeting it are equally unethical, so dismantling capitalism isn't the answer because unethical people will just take their greed off the balance sheet and stuff it in to warehouses and gulags.

If ethics were society's most pressing problem, we'd be having a hard time finding things that are wrong. We're nowhere near running out of moral failures. Would that we could power the grid with them. Maybe we can, but somebody got paid to say otherwise.

My take is that bolded part is a valuable rule of thumb for telling us when we need to work on ethics rather than moral failures.

Moving on, the latest ethical drama is the present generation of chatbot AI which is presented as some ridiculous threat: helping students create fake papers, criminals plot crimes, scamsters scam, the insane commit suicide, and the PHB be idiots (to name a few recent concerns). No serious moral failures have come up. These are all things that could be a problem, maybe, and when they're illegal or against rules, would stay that way.

Meanwhile we're up to our eyeballs in all kinds of crimes, scams, frauds, and villainy that just isn't that hard to sort out ethically - and certainly not AI-based. We don't have a problem finding things that are wrong. That tells me that we have much bigger ethical problems than AI. And looks to me like how we address that doesn't change or improve no matter what we do to AI research.

So when I see a petition like the letter in AnonTechie's journal that demands a six month pause in AI development globally, I am exasperated. If we really followed through on that letter fully and honestly, how would we have progressed even a little on the problem? We still wouldn't have or understand advanced AI. We still wouldn't have any idea how to fix its problems. We would have just wasted six months of valuable research time and peoples' lives and be back to square one - making the case for yet another six month delay because nothing changed.

Edit: Is it time yet for a Trump update journal? Seems there's several active court cases surrounding him now.

In a thorough and well-reasoned decision, Judge Terry A. Doughty of the U.S. District Court for the Western District of Louisiana has denied government defendants’ motion to dismiss in State of Missouri, et al. v. Joseph R. Biden, Jr., et al. The New Civil Liberties Alliance, a nonpartisan, nonprofit civil rights group, represents renowned epidemiologists Drs. Jay Bhattacharya and Martin Kulldorff, as well as Dr. Aaron Kheriaty and Ms. Jill Hines, in a lawsuit that has exposed an elaborate, multi-agency federal government censorship regime. Judge Doughty wrote, “The Court finds that the Complaint alleges significant encouragement and coercion that converts the otherwise private conduct of censorship on social media platforms into state action, and is unpersuaded by Defendants’ arguments to the contrary.”

Discovery in the lawsuit unequivocally establishes that at least eleven federal agencies and sub-agencies, including CDC and DHS, directed social media companies to censor viewpoints that conflict with the federal government’s messaging on topics ranging from Covid-19 to elections. Federal officials engaged in a lawless, expansive censorship campaign that employed illicit tactics—including coercion, collusion and coordination—on social media companies to suppress the airing of disfavored perspectives on Covid-19 and other topics. As a direct result of state action, NCLA’s clients were blacklisted, shadow-banned, de-boosted, throttled, and censored, merely for articulating views opposed to government-approved views on Covid-19 restrictions and regulations. Judge Doughty held that “Plaintiffs have plausibly alleged state action under the theories of joint participation, entwinement, and the combining of factors such as subsidization, authorization, and encouragement.”

In confirming Plaintiffs’ standing, Judge Doughty said, “The threat of future censorship is substantial, and the history of past censorship is strong evidence that the threat of further censorship is not illusory or merely speculative.” Judge Doughty also found Plaintiffs’ alleged injuries-in-fact are “redressable by the Court,” and that Plaintiffs had demonstrated sovereign immunity does not bar their First Amendment, ultra vires, or APA claims.

In case one thinks this could somehow not be abused, I'll note a year old study that alleged over 600 times where a large social media outlet (here called "Big Tech") suppressed speech in favor of Joe Biden's 2020 campaign or subsequent administration in a two year period.

MRC [Media Research Center] Free Speech America tallied 646 cases in its CensorTrack database of pro-Biden censorship between March 10, 2020, and March 10, 2022. The tally included cases from Biden’s presidential candidacy to the present day.

The worst cases of censorship involved platforms targeting anyone who dared to speak about any subject related to the New York Post bombshell Hunter Biden story. The Post investigated Hunter Biden and the Biden family’s allegedly corrupt foreign business dealings. Big Tech’s cancellation of that story helped shift the 2020 election in Biden’s favor. Twitter locked the Post’s account for 17 days. In addition, Twitter slapped a “warning label” on the GOP House Judiciary Committee’s website for linking to the Post story.

Big Tech even axed those who blamed the current inflation crisis on Biden. For example, Facebook censored Heritage Action, the advocacy arm of the conservative Heritage Foundation, on March 15, simply for posting a video quoting Biden’s embarrassing statements on energy policy. Facebook placed an interstitial, or filter, over Heritage Action’s video, suppressing the post’s reach. The video showed Biden and officials in his administration explaining how his policies would cause gas prices to rise.

But the largest category by far included users who dared to call out Biden's notoriously creepy, touchy-feely behavior around women and children. The 232 cases of comedic memes, videos, or generic posts about Biden’s conduct composed more than one-third of CensorTrack's total instances of users censored for criticizing the president.

The allegations of the lawsuit are interesting. For example, two US states (Missouri and Louisiana) are among the large mix of plaintiffs and there are assertions that they have special standing to sue for various reasons (such as protecting the interests of their citizens and quasi-sovereignty status). A few of the claims are: various acts of censorship described in the suit are unconstitutional because they were carried out at the behest of the US federal government - censorship by proxy; that in addition, the two states have a duty (based on their state constitutions) to protect the speech of their citizens (I believe every state in the US has something similar in their constitutions); and that similarly, the two states can't hear all the voices of their citizens via social media because of this selective censorship. While the last is a dubious legal argument in this situation IMHO it does bring up an interesting point. If a future state government relies on the large social media platforms to be informed about what its citizens want, is that something that is legally actionable on anyone? Obviously, there's a variety of deception possible with fake accounts (possibly driven by AI) and such to twist the perception, but some of that exists even with older technology like mail. The practice of astroturfing, or creating a fake "grass roots" lobbying campaign, started as automated mass mailings and phone calls to legislators.

So I guess the TL;DR is that once again the courts are addressing abuses of censorship by proxy along with the problems of how a government communicates with its citizens in an age where fake communication is getting easier and real communication can be censored most likely legally.

Note that Civ and other strategy games in the same ilk got it wrong.

Science/Research is not something you "bank". Science is something you invest in, and continue to invest in. Stop the investment and "Science" regresses.

For those who haven't played the Civilization games, the idea is that you run a civilization and compete with other civilizations. They typically are city-oriented (a key step in expanding power of the civilization is building more cities), and they invariably bank science. That is, if I'm researching writing and decide for some reason to switch over to research other technologies for a while, I can pick up where I left off on writing - even if it is a thousand years later!

This is part of a larger problem, namely that civilization is seen as a strictly progressive affair - anything you do aside from losing wars moves the civilization forward. Sure, if you start next to Alexander the Great or nuclear Gandhi (or worst of all a human player!), you have plenty of opportunity to experience civilization setbacks as the aggressive neighbor makes war on you. But you can slack off on science, infrastructure maintenance, etc and pick it up later.

In the real world, there's plenty of failure modes other than losing wars. Conversely, a number of real world civilizations have lost a bunch of wars yet still were able to keep relevant.

Over the past couple of weeks, I've been watching a podcast series called "Fall of Civilizations". Production values do leave something to be desired (such as showing a trash fire in an unnamed Middle East neighborhood when discussing someone burning something in a war or rebellion or reusing stock images in multiple episodes), but it's an interesting angle on history. The author starts with a discussion of the significant ruins that the civilization left behind, often from the point of view of historical figures who discovered it first, what the civilization actually did that made it notable, and finally, what led to the fall and its aftermath - including from the point of view of the people caught up in the fall. Last I checked he's up to 16 episodes.

What's interesting is how few of these disasters have a single, clearly identifiable cause (sometimes they just don't know why at all). Usually it's multiple factors with considerable uncertainty as to the relative significance of the factors. Again, this isn't something captured in historical games like Civilization. For example, what combination of factors caused the collapse of the Assyrian empire (Episode 13)? Did it fall due to the fact nobody liked them and finally unified enough, climate change transition from the best rainfall in the region to something of a megadrought, vast overextension of the empire (key parts of the army couldn't return in time to the core Assyrian region to save it), or an effect nobody has considered yet (maybe some sort of heavy metal poisoning explains their wacky leaders)?

A key problem for many of these civilizations was that they either stopped banking something (the collapse of the trade networks and disappearance of multiple languages of the late Bronze Age cultures of the eastern Mediterranean which at least partially was due to not diversifying critical resource needs like bronze and food and very low literacy) or they had some hidden deficit in their civilization that grew over time (such as the decline of Sumeria due to widespread salinization of irrigated farmland which when the region was hit with a drier climate turned the area from a great net food producer to mass starvation).

That leads to my observation - that just because a society or civilization has something now, not just nuclear engineering know-how and experience, doesn't mean it can keep it. Too often strengths of civilizations are taken for granted and just assumed that they will continue no matter how much we or the environment impair them. Well, there are a bunch of dead civilizations that indicate otherwise. You or nature can break something to the point that your civilization no longer exists in a recognizable form.