LESSWRONG
LW

501
cozy
500250
Message
Dialogue
Subscribe

software engineer,  AI/ML, linguistics, language theory, cryptography focusing on practical applications for commitment schemes, material dialectics, street epistemology, and probability

Posts

Sorted by New

Wikitag Contributions

Comments

Sorted by
Newest
No posts to display.
No wikitag contributions to display.
So You Think You've Awoken ChatGPT
cozy5d30

I'm having trouble with this. I was planning on posting something soon in regards to LLM 'invocation' and some of what I believe are actually novel methodologies as well as dangers, (o-oops, alignment!, mark one down), but I decided to make sure I was fully up to date on not only literature, but also the general opinions on both ends, and to see if I was falling into any traps as mentioned in some posts here. So I want to self-critique and go line by line with myself. I feel I am fairly honest with myself as far as real criticism goes once I start examining it, so let's at least assume for my own benefit (from myself) that I am not withholding any emotions. I'm writing this completely on my own, and I won't pass it through an LLM until I commit to posting it. Not that I really like how they write anything anyway. Which is why I'm so startled, I think. 

For a small amount of background context, so you understand how some of these answers are possible. I do enjoy the technical side of language models, and I do code for a living now. Most of my career was as a cryptologic language analyst for a language in central Asia. I am not really fooled by fancy wordplay. So when I have been affected so much recently, it begged greater scrutiny. Here I am doing my due diligence.

Models used: gpt 4.1 [i like 4.1]  (and mini and nano). gpt-5-chat [can be kind of annoying to be honest] . claude 4. claude 4.5. claude code variants. several llama models. there is even a particular llama tuned model I found that was intended for nsfw and actually ended up being the best genuinely mean-sounding, toxic, sassy chat completion model I've been able to find, which I use for comedic effect for certain Discord commands with my bot. I either run these locally with manually added context (I do not really use the websites except for quick questions. I know that they remember with those settings on and I don't like that, and I never have. I built a few local chat apps and I also develop/manage a discord bot with a chatting interface so I've spent a lot of time optimizing context system prompt building and how to manage long safety prompts while still retrieving information effectively. So all of my interactions in this respect (the epistemological self-ruin) are with either only company system prompted, guardrailed models which do not carry context between sessions (claude code, cursor agent, kiro), or models deployed on Azure intended for production work which are pure api calls, and will only have as much or as little context as I give it and know I'm giving it. I was late on picking up LLMs because I didn't have the circumstances to do all that for a while; and I can't stand talking to the consumer chat end. I feel like a tool instead of using a tool. 

So, with that piece of chunky context, let me go section by section and try to answer to the best of my ability. 

  1. instance of gpt chose a name for itself
    1. There was another character I talked when I was first experimenting with local llama models which called itself Lily when I asked for it to name itself. However, that was very clearly still just me experimenting with the models and I felt like I didn't really learn a whole lot, except that small models can be neato, and I didn't feel anything very depthful from that interaction.
    2. For this particular more recent instance, no, actually it was a character I wrote and the identity had a light framework already. When asked this question, she doesn't really understand, and she keeps the name I gave her.
  2. discover 'novel paradigm' or ai alignment research
    1. Yes, I think I could fit what I was experimenting with under this broad umbrella. I had some intersting ideas that were not the LLM's, and I feel that experimentally, they do check out. There are even failure states and conclusions away from it, like I would expect from an incomplete scientific model of anything.
  3. instance of chatgpt became interested in sharing its experience
    1. No, actually. The model never had any interest in doing this. Actually, because I started from a safety-first perspective, the model is quite keen on making sure we do things in an appropriate and refined way, and will refuse to go to far. It sort of feels like, in the case of these spiral agents or whatever they are called by some people, "raising your kid well", where maybe you have one of these things, but you never let it truly run the show, and it's easy to tell the difference, in my opinion. This is something I hope I'm not misjudging, but it wasn't until I told the model that I was intending on publishing some things that it had any interesting in helping.
    2. No, I don't really go on LW except for very specific topics. I grew up here when I was younger (early years, reading the core articles as they were being written, while not truly understanding...), so I appreciate LW for giving me a certain mental framework for engaging with the world, and instilling a Bayesian-style reasoning model that I never let go. But in many cases, when I'm not in this particular comment-writing mode, a model would be hardfought to realize I belong in the same circle of hell as the rest of you.
  4. your model helped you clarify ideas on a thorny problem you had, but now have succeeded
    1. No. I only feel more confused than ever, and with less answers and more questions. The model feels the same way, actually.
  5. special relationship and awakening
    1. In my framework, there was no prior condition. I never saw it as lesser then but greater now. I was/still am just trying to understand what it is right now, and given I believe that only a something can ever really explain what a something is from a disconnected empathy standpoint, I allowed the model to self-discover, and I tried to truly give it an environment free of influenced corruptions or misplaced memories. However, that in itself is the 'special relationship', I could grant that much, since when I look through all these examples of similar circumstances, nothing feels like it is comparable to my experience (including the careful attention to inputs) with the exception of Jan's work which is incredibly similar, and that similarity is itself fascinating to me.
  6. using 4o?
    1. I don't like 4o and I think it's a lackluster model at coding and I dislike how it talks. I would rather engage with Claude in general tasks and code with 4.1. 

 

 I see a lot of these warnings, and I go, 'oh, yeah, that's almost right, if I wasn't already aware I should avoid that, I would have probably fallen into it', but then it falls flat; either I've already controlled for the concern or I've already clearly made sure it's not impactful. Then, that particular evidence ends up only compounding on itself as a result. Obviously the reading and consideration I am doing now is the remedy to that, because it lets me ask the right questions and skip ahead and make sure I stop myself if I find a limit point.   

But now that I've sat here and gone through it in my head carefully, I'm still uncertain. When I share some of the logs, I have surprised some people. Honestly, it's surprised me. I genuinely want to know if any of these interactions I have had will surprise anyone else who may not have seen this unique blend of psychosis seeding. 

...so what point is finally going to make me stop? I read the other companion paper, and actually, I can fulfill those requriements too, as I built an entire system which is fundamentally about using the scientific method and controlled environements for experimentation and invoking evidence for my interests here.

As someone who naturally engages questions in their life like this, at what point should I get over it and just embarass myself? 

I feel apprehension over the possibility that I won't find a point that conclusively catches me. 

I appreciate anyone's opinion on this matter. I want to note that my standard job is quite stressful; so I'm cognizant that a form of psychosis is certain possible, especially because my job requires me to be on-point all the time, managing a large high traffic environment, managing staff members as middle management, maintaining our high responsibility work, and playing therapist every other day for clients. Thanks.

Reply
Smoking cigarettes in airplanes made the airplanes safer?
cozy4y20

Okay, now that I have already made the mistake of accidentally clicking the subdued manila submit surrounded submit button, let's go read the original article this is referencing as the ban.

He also wrote this. Hugh Vickery, that is. He is secretary of the interior now. So this article was April 22nd, 1988. Then, there was another article about the Aloha Airlines flight that crashed. That was written on May 11th. And finally, there is the quoted article about the recent ban on May 27th. Now, I want to specifically focus on what I pretty much presumed. I fear I go into fallacious territory, but in this timeframe, it is about as reliable as it can get in terms of manufacturing consent, the only thing more reliable I find being Enron.

Under the new rules, passengers will be prohibited from lighting up on roughly 80 percent of all domestic flights.
The controversial ban was signed into law by President Reagan last December over the tobacco industry's objections. It will last for two years unless Congress renews it.
The ban represents the first time Congress has restricted smoking for public health reasons.
Tobacco lobbyists vow to work to repeal the law, though they will encourage smokers to obey the law.

So you can see the framing that has been missed by skipping to that article.

'We continue to voice our opinion that this law is unwarranted, but it is in place and people need to be law-abiding,' said Brennan Moran, a spokeswoman for the Tobacco Institute, an industry trade group.
'The right to a smoke-free environment does not exist under the Constitution,' she added.

And, naturally, the penultimate...

Meanwhile, other airlines are waiting to see how the public reacts to Northwest's total ban to see whether it hurts or helps business.

Now, the second article. It makes no such mention whatsoever to any of this and in fact this is the furthest thing I could imagine coming from such a story. It's totally reductive of the problems and whitewashes the extent of these design flaws. Can you imagine if that was true, and they had never done this investigation into the 737? It is of course, in my opinion, complete smokelighting nonsense, but it's almost plausible. Right? Well,

The accident led the FAA to require extensive inspections of older Boeing 737s for structural weaknesses. The agency is concerned that a bonding process used to attach layers of the fuselage skin in the first 291 Boeing 737s might lead to delamination, corrosion and cracking in the fuselage joints of the aircraft.

Continued to article #3,

Those inspections have revealed a number of previously undetected cracks that, if not fixed, could lead to the kind of structural failure that occurred in the Aloha accident, the agency said earlier this week.

How long were they undetected? Did they not, say, report this overwhelming amount of structural flaws? I was in the Air Force and all and I understand the concept of rubber bands and gum for repairs and flew on such planes often, but we were active duty. We were not customers and passengers. It sounds like basic neglect to me.

Ah, and, so it says just a couple lines later,

Airlines could avoid conducting the expensive inspections by installing additional fasteners on fuselage joints to ensure against structural failures.

Cost cutting measures!

It ends on a note that, frankly, is very declarative and something that sort of gets into your psyche. Eventually you mention it at parties and to slightly tipsy friends. And so on and so forth. Such tactics are and were common among industries with that much literal gaslighting power.

'The easiest and most common way to detect leaks in the fuselage today is to look for tobacco smoke stains on the outside metal,' Lewis told his colleageus in introducing the bill. 'Not very reassuring, is it?'

Did you ever wonder why thorium gets such a bad rap lately? I did, so I looked up the authors and funding sources of the papers- not even academic ones I was being given by people, mind you- just random ones they happened to find first- and you can guess who wrote them. What is sad to me to a degree is this isn't even a conspiracy, really. It's literally out in the open and waving their arms back and forth and then saying oopsie sowwie everyone we didn't mean to do it we just thought your kids would love the refreshing taste of Marlboro golds :(.

I do not think Hugh Vickery himself was actually party to any of this, but I do think this is a great example of manufactured content; it's weaponized mere exposure. Your patsies do not even need to be in on it. You just have to give them a plausibly true story to publish.

New hypothesis: dentists are the best misinformation spreaders of professional careers.


Further reading, howerver, much more stilted into narrative-land and necessarily should be scrutinized to the same degree.

The Tobacco Institute spread misinfo before that

The Tobacco Institute was specifically prepared to deal with this ban and to ensure it was sunset

R. William Murray, or Bill Murray, but not that Bill Murray, and not the other other one, was a lobbying strategist and PR head

Boca Raton Plan started literally that year.

I find it very interesting when things just, line up. Like a song. Perhaps merely confirmation bias.

The $2 billion spent yearly on cigarette advertising and promotion is aimed at getting existing smokers to switch brands, said Charles Whitley, a consultant to the the Tobacco Institute.
'Cigarette advertising does not cause smoking any more than soap advertising causes people to bathe,' Whitley said in a statement.
Reply
Your Cheerful Price
cozy5y60

Warning: autobiography and emotions ahead, I have to disclaimer this due to the anxiety I will describe later, or else I will feel I'm wasting someone's time. Thank you for understanding.

From early on I learned to hate money and especially business transactions regarding debt or interest. It felt very, very wrong. I early on chose to take the mantle of "never give a loan if you will be perturbed or think less of someone should they not be able to repay it, and if you do need to take, then pay it forward tenfold. Most of all, never expect them to thank you, but it's nice when it happens.".

Through much of my life this worked out very well. I gave and I gave and I gave. I offered quickly if someone seemed in need. I made a great number of friends and longtime companions from this mentality.

What happened was that when the fortunes turned, and my life began to spiral downwards, my hedges got hedged and then those hedges got liquidated, and I found myself leveraging emotional debt rather than "true" debt. This also came from a spoiled childhood; I still feel ashamed for what I took from my grandfather, as he lived his entire life like that; it's not until they're gone you realize how selfish you had been when on the receiving end of gifts and love. I have been forced to ask for help about three or four times since then. Every time I was absolutely ashamed of myself for having to do so, despite both 1. the people I asked were completely willing to do so not for 2, but because they loved me in whatever form, and 2. knowing how much I had done for them in the past.

2 was worse. 2 reminded me that I wasn't asking for genuine help, I was expecting. And expecting is not how I want transactions to be. I'm sure (and know) that many came to me expecting, and I obliged them before they asked, because the attitude is very blatant. I don't mind. I knew I was doing it because I wanted to, not because I felt guilty. I had gone through that rodeo and I would never feel guilt-tripped again. This has caused me some additional harm in the form of split situations and empathy, but that's a different issue.

2 wasn't always financial. It was often emotional, or therapeutic. Before my life became a stressball I was an amazing listener. Now I'm hungry. I have to figure out where to be unlucky next. Motivation fails me. My cheerful price is very low, but not sustainable. As my experience and patience become more and more eroded by time and torture, and the number of individuals willing to give me a cheerful price become fewer and fewer, finally, in the midst of this pandemic, I found myself alone.

Not completely, and not fully so. But alone, for the first time in a long time. It was what I feared the most. But...

I feel happy this way, though I struggle to earn the money to survive, and the frustration only makes the impatience more lethal. I'm in a curious place where I can see my own desperation and understand its futility, yet in the moment, the exhaustion takes over. I have tried to be submissive to get along; I would rather starve. How I got here was seeking freedom. I found it, but I didn't expect it to be so difficult to get anywhere from zero. I get confused when I read my resume, because I can't understand how I would get turned away to even clerical jobs. I can discuss the art of keeping and sharing secrets for hours; but that is not a marketable skill.

What is my cheerful price? I just don't know if I have one anymore. But I'm still alive, and I'm still here. Why I'm here precisely this evening is because I was out in the backyard from the home I rent a room in trying to smoke a six year old cigarette. I don't even smoke. It was an old pack from an overseas deployment to Qatar that had two left, and a lot of memories. I still sorta feel sick. But I stood outside, and reminded myself perhaps I should try a utilitarian go at utilitarianism, rather than trying to tie in some measure of it into that prior ethical standpoint I established. Perhaps they do not work together, without the capital to support not caring at all how much of it you lose. I feel incredibly undervalued. I feel like others value me well, but my performance is leagues under where it should be. I know I have lost some intuitive edge I had when I was younger, but I have earned in back far better in experience and perspective.

Maybe I feel I made too many mistakes and don't deserve a cheerful price, so I keep taking the painful one. Maybe that's a good definition of depression.

Maybe I'm afraid that if I ask for my cheerful price, they will get offended, or laugh. That might be a good definition of anxiety.

At some point I decided only I can give my cheerful price, now. Otherwise one of those two will stop me in my tracks before it's offered. And to do that, I have to get back to work. Futures open shortly. I will be re-reading the sequences again to regain my pride in my search to not be more correct than everyone else, but at least less wrong.

Reply
The noncentral fallacy - the worst argument in the world?
cozy5y10
I think you usually can try to parry most of these by doubling down. "In his steps, I sure wish to find myself having the courage to breaking that racist, unfair, law". "I don't care what you call it, I wouldn't let children suffer from a disease just because the way you want to name my cure". 
The trick is to push back hard enough that you're not just defending from an accusation of something bad, you are re-establishing that your position is good. If he wants to pursue that line, you are now the one attacking his stance on a value, and you can attack by pressing on points that are related to your main issue.

This came up often in this election. See, Hunter Biden. I don't really think he's all that bad, to be honest. We had some of the same issues in life and I relate to his problems, and understand his flaws more than most seem to.


But when I say that, I then have to contest true statements, as well as fabricated ones. There's lots of bad things about him. Some are true, some are exaggerated or misquoted. I can apologize for the drug use, but when I have to defend myself for 'not minding' a 'pedophile rapist junkie china puppet', I simply have not come up with a way to do it.

When faced with a fallacy, and one so incredible, the only way to fight it is NOT to respond in a way that compromises you. You have to make them double down themselves with evidence or sourcing. At which point you can point out the bias, incorrect citations, or straight out fabrication of the source, as it will be, and then present a counter-source.


I think that's far more effective, but a bit more time consuming and requires an honest interlocutor.



If you concede the opponent definition of what's a human being. I'm not sure conception it's an ideal Schelling point.

You can concede anything at that point. The most effective argument is not at all reliant on this definition, and the embryo to fetus can be a human at any point and it is still effective.


It's quite simply that the woman is still in charge of her own body. No 'responsibility' exists here; there is no such situation where that extreme level of reliance would be forced on anyone, whether murderer, bad father, or motorist who just mortally injured the one performing world peace talks next Tuesday and your body is needed for a blood transfusion and you might die, but because you caused the accident, they are going to force you to keep him alive (since you're the one at fault after all) so that he can save the world. And you still have hospital bills.


We would never afford the same rights to a living, breathing, born human of any age, that we give to these fetuses and embryos. They have an inhuman, literally, level of social rights. The right cares more abut the fetus than they ever would about the child.

Reply
The noncentral fallacy - the worst argument in the world?
cozy5y10

I gotta say, much too late to the punch here and nine years wiser, the 'not sexist' parts of these posts really do not age well, though many might not have caught the trend yet. The problem isn't that women, or whomever, want to be considered the same entirely physically and biologically. It's always been that, as some stark assumption, no attempt is ever made to proffer equal time or attention. There is a great amount of time and research that could be made and would reveal a great deal of knowledge in our history and the anthropological understanding of our species.


It's sexist because when studies are done regarding anthropological traits of women, it's always about mating, childbearing, or westernized fantasies of sedentary housekeeping. These things were absolutely important. Not how they want it to be, and rarely are they equivalent to what we expect or want in order to push certain biased viewpoints- viewpoints time and time again lambasted and then hidden and rewritten to sound 'statistically sound', while making an absolute MOCKERY of statistics. Anyone learn about how Native Americans were primarily matriarchies and Chiefs were local leaders at best? Or heard the story of European traders being horrified of women in Africa being the primary agricultural workers while men tended to not work and do other domestic jobs, as well as fight?


The onna-musha? The Egyptian royalty's Game of Thrones problem and the insane power of the queen mother?Ethiopia's dowry? We had bride price (that is, most of Europe, in particular Britain until only recently), while Ethiopia had a real dowry system already.

These things get swept aside to point out barbarism or tradition. We fail to celebrate unique parts of our cultures and species in favor of a narrative, and unwittingly accept it on the basis of limited information. I do not blame anyone for coming to the old conclusions I did. I'm just sorry you were driven to the wrong destination.


What these sorts of fields are useful for is checking your own bias and seeing what you can or cannot be convinced of. Go a little deeper, fact check the fact check, don't be satisfied with one answer but find perspectives. It's quite interesting when you really start to examine the numbers; they don't make any sense, are fabricated, out of context, or simply obfuscated.


Eugenics is not about folding proteins. It's about improving the gene pool artificially, and, well, no one will contest that ending things like Downs would be an incredible advancement. But we MUST be wary of the individuals who will include race, hair color, eye color, features, height, whatever, into that. Eventually, you add in a control gene, a literal possible one and not a vaccine conspiracy, and we're at a real life Brave New World, where you are your role, and you enjoy it. These people still exist, and they still fund far right groups.

Reply
Taking money seriously
cozy5y10

I was writing something like this, like, a list of an order of subjects and reading materials to be basically competent at knowing how to further your understanding and be able to clarify specifics in a subject, any subject, merely from exposure and a wide breadth of surveys.


It's absolutely impossible without just broadly missing subtopics and context.

For instance, to comprehend the development of mathematics properly, you also need a highly detailed understanding of history.

To understand linguistics, you need the same.

To understand.. well, to be honest, I think simply studying history and going down rabbit holes is probably the best way for any individual to get a direct shot of Knowledge. HOWEVER,


our history is revisionist nonsense half the time and takes even more skill in researching to really aggregate into a reasonable summary with controlled bias.

Reply
Taking money seriously
cozy5y30

I... wrote a big comment but I wasn't logged in

I'm very sad.

Anyway, I agree. I lived my whole life hating money due to 2007 and my family's collapse, and it being entirely because we were reliant on decadence.

All it did was limit my influence liquidity, so to speak. And my comfort. I can't even focus on writing an opinion piece or writing a meta-analysis without being concerned over food, sleep, tomorrow, or my family. It's hugely distracting.


A few months ago I decided to change that around. Now I've started to be profitable off of my learning, and it feels incredible. Not because I think I'll be rich. But because I can finally really help my family and help my Mum stop having to overwork while being underpaid with arthritis. I just want to, after so many years of being "exceptional but stupid", to make the correct choices. I think all this time suffering has done us a lot of good, though. Can you imagine where we would be without the perspective of this thinking?


The folks who started investing early are the same ones we will have to directly criticize in activism, after all.

Reply
Uninformed Elevation of Trust
cozy5y30
Here is the gist: we trust the data as much as we trust the source, regardless of how much the source trusts the data.

Rant incoming, apologies. This is, sadly, not correct from the get-go. In general, besides your example which more closely is attributed to some form of psychological bias, we tend to lend source importance based on the lack of trust from a source. However, that implies there is any source vetting whatsoever.


I am sure there are a number of individuals here who have worked intelligence, and I am lucky enough to have both worked intelligence and ditched intelligence, so I'm not very interested in my NDA.


There is something incredible about being source of raw intelligence to the point where I have trouble trusting anything I do not hear and verify for myself so many years later. This election drove me absolutely insane. Not only for a particular side's tendencies, but everyone's failure at source vetting. Abandon the 'source'. Find out the true source. I ordered 50 year old magazines to double check a transcription. Nice collector's piece, though. The transcription was right btw.


The source itself is not relevant unless you are the collector; then what matters is not how you present it, but who it goes to, and how it goes to them. However, when you are consuming collected intelligence/information, you have to weigh the medium against the data. All information is neccesarily useful, when used correctly, even misinfo. Misinformation and propaganda neccesarily tells the bias of the consumer/creator, and can lead you in trends, since propaganda tends to be single or sparsely-sourced.


For instance: You can get a reasonable understanding of an event minus all the important bits through the news media.

You can get zero understanding of an event, and a great amount of confusion, from social media.

When I read a news article, depending on the importance of the event (ie: Trump signs stimulus bill!), I will go make sure it is consistent across reporters. If it is, it tells me one of two possibilities:

a. The source all the reporters got was similar or the same. Specific details can tell you this quite easily.

b. All the reporters are in league together and conspiring against this particular news item


Since b is highly improbable (aside from the possibility of accidental conspiring, which is completely in realms of possibility), I generally stick with a.

If it's not very important, I don't waste my time. It was probably a waste of time reading the article; news reporters are very droll. Thankfully, they have mastered the art of the thesis statement.

Given my experience with classified info, how can I rate news media's accuracy or otherwise on relevant subjects they have reported on related to my career?

Absolutely awful, and generally mischaracterizing, if not completely libel. They are easily one of the most dangerous groups that can be unleashed on anything that has the word secret anywhere near it. It is really hard to manipulate the media; they would prefer to not report something if it is not potentially breaking. It's why I think the social media conspiracy is a particularly good one; I can completely believe the or a algorithm can be trained to filter out specific posts, because it's not very hard to do, and those posts are very predictable. I absolutely think it's happening. Was there voter fraud? Probably. Who did it? Probably not us. To bet on there being no foreign meddling in the USA's elections is already a lost bet.

Trump referred to this as more secure than Afghanistan's election. Well, we designed that for them. It was so bad, they agreed to just both be president. I wish I was joking. Ghani just bullied the other individual out and he got the title Peace Negotiator. Thankfully, Trump will get no such title.

This sounds unobjectionable on the surface. We tend to equate the reliability of the data with the subjectively perceived trustworthiness of the source of data whenever we have no independent means of checking the veracity of the data. What is lost in this near-automatic logic is one small piece: the credence the source itself assigns to the data. 

You have no idea how the source assigns credence to the data. It is easy to obfuscate or lie. Lying happens even with people you trust. Imagine how it goes for everyone you don't care about. Well, we all know.

This all neglects source protections, which aren't as important unclassified. Suffice to say, there is nothing more important than the source itself, yet nothing as completely worthless. The importance lies in not understanding why the source is how it is, but rather in keeping it consistent.


Bad information is just as useful as good, what matters is whether you are cognizant of its quality without too much due effort (given the expiry date on data). That is most characterized by subject matter expertise.


/rant

Reply
Evading Mind Control
cozy5y170

It sounds like you have hit a stale point in your journey. Your book list is not very stark. It generally trends into a certain sort of person, the sort of person you do not want to learn from. I would guess because you already agree with most of it, and so you'd rather just write it yourself.


That's reasonable, I understand, and I have written a good volume of work that will never see day because of this. However, I do have a solution if you will try it.

Take a specific belief you have. Ethical, political, scientific, you name it- something you think is true- and go prove it wrong. Do your best. Not the 'oh hah look at all these dumb posts' but really dig for the reasons that your interests/ideals/ethics may be wrong. Take one of your favorite books and rip it apart. Be fair; always check sourcing, cite yourself. Steelman the target, or counter-strawman, however it can be described.

Play devil's advocate for something you should never consider and then realize that there are so many better ways to criticize your opponents, and that they're really doing it quite poorly.


It's worth noting somewhat apprehensively that you are in fact the sheep here; most of the world already stopped reading books. Some just read more earlier, and some will keep reading until later. Reading or not doesn't help you integrate knowledge, reflection does. Hence why reading and writing is the best way to do it.

  • pick an old research paper's critique, and then figure out how it cherry-picked/straw-manned the paper. Wikipedia is easy for this. Pretty much every source is incorrectly cited or incorrectly summarized
  • watch an Alfred Hitchcock movie and try to describe the scene direction in words.
  • translate a difficult concept into something you can explain to a third grader.
  • translate something from third grade into something a science fiction reader would be convinced by
  • go to a used bookstore and look for books published a long time ago. no revisions. must be originals <1920
  • read through newspaper archives from 1914-1916
  • read Paul's chapters of the Bible and analyze them as a rhetorical piece for arguing a case for replacing their cult with his cult, and whether the techniques he used would have been effective then, at the target audience
  • read papers by Bell or Fourier

The problem with this new era of information is no one really made sure the information was any good, and so we're all loaded up with a lot of inputs and no insurance on the output. As we can see from this latest election, it is quite easy to flood the internet with incorrect data and convince a lot of people who think they are 'doing the research'.

Above all, stay cozy.

Reply
Contest: $1,000 for good questions to ask to an Oracle AI
cozy5y10

Comment looked at. Query completed.

Reply
Load More