While the post didn't particularly speak to me, I think you are being overly reductionist. You can dismissively summarize anything into a few "key points."
Just in the first couple of paragraphs alone:
- The author discusses the "why" by pointing to research on how new information triggers dopamine pathways. Maybe this is obvious to you but it's not necessarily obvious to everyone.
- The author points out that this would have historically been an evolutionary advantage (at least that's my interpretation), but access to modern high-volume, low-nutrient information has made this an addictive unhealthy habit.
- For those already familiar with how modern access to high-calorie, low-nutrient food (that triggers the dopamine response), the author is showing that the same mechanisms are at play with information.
I think it's actually your comment that contains the tweet-able one liners (as "key points") that are likely to result in head nods and smug self assurances but aren't doing anything to encourage deeper thought.
Honestly, I like both of your posts. Thanks for the follow-up. Your thoughts are insightful. I hear it a lot that "too much information is ruining us", but what about crap daily newspapers that came in mass around 1900, then radio, then over-air television, then cable television? All of these are about providing large amounts of low quality information. These days, you can get satellite TV with 500 (five hundred!) channels. It must be 99% rubbish content. I guess the difference now is that the web is much more interactive, and our minds can be more easily hooked as a result. Dunno; when I was growing up, the kids who gorged themselves on shitty cable TV weren't the brightest bulbs in the chandelier. I feel the same now in the Internet era.
My point for this post: Each time I hear (paraphrased): "It is different this time.", I try to imagine the world 25 years ago. Then think: Is this really
that different? Most of time, I think, "No, it isn't that different." (I use the same mental exercise when I hear people say that "this generation is so different than the last for reason A, B, or C.")
> but what about crap daily newspapers that came in mass around 1900, then radio, then over-air television, then cable television?
a) None of these were, first algorithmically, and then using machine learning, optimized to constantly grab attention and maximize interaction time on an individual basis.
b) None of them had the interactive effects where peoples interactions with one another are guided and used, to draw them into echo chambers which are, again, designed to maximize interaction time.
> Then think: Is this really that different?
Yes. It is. Simple example: How do politically motivated troll armies influence an election in a reality where people don't use social media as their primary news source? Answer: They don't. They can (and likely have) bought some space in some low-quality tabloid through strawmen, but that has nowhere near the range and impact of one guy in some government office pretending to be 100000 "Average Joes" halfway across the globe.
> How do politically motivated troll armies influence an election in a reality where people don't use social media as their primary news source? Answer: They don't.
To what degree do they "influence" an election in our "reality"? (And, how many people can even genuinely and substantially wonder what the truth of the matter is after growing up in countries subject to Western "influence"?)
In percentage terms please, and please explicitly acknowledge opinions as such.
Sorry for ruining a good story (did you even notice you didn't answer the question, but instead posted a bunch of stories...how do you think these things work), but the topic and specific nature of the accusations (misinformation) demands it.
the articles they posted had lots of explanations of how the writer of said articles thought things work, perhaps they (the poster) thinks that these things work as laid out in the articles and thus are not required to explain in extra detail how they think these things work.
In relation to percentages, lots of the articles had numbers, you could also do the work of deriving percentages from the numbers if you were so inclined.
>did you even notice you didn't answer the question,
the way English and Internet communication work if you quote a question and then give a bunch of links it is reasonable to assume that is the way they are answering the question - in short they believe that the links they provide are a good faith explanation of 'what degree "they" "influence" an election in our "reality"'.
>and please explicitly acknowledge opinions as such.
You seem to want them to do an awful lot of work to answer your short one paragraph question!
> the articles they posted had lots of explanations of how the writer of said articles thought things work, perhaps they (the poster) thinks that these things work as laid out in the articles and thus are not required to explain in extra detail how they think these things work.
My question was whether the poster realized propaganda/misinformation ran on stories, since that (and only that) is what he was posting (with no accompanying assertion, a genuinely impressive technique, if intentional (no accusation, just sayin')).
> In relation to percentages, lots of the articles had numbers, you could also do the work of deriving percentages from the numbers if you were so inclined.
You could also get some percentages with a random number generator. Are you asserting that reasonably accurate quantitative (percentage) truth can be derived from these articles? If so, I'd like to see you explain how, and also how you would determine your theory is correct in fact.
> the way English and Internet communication work if you quote a question and then give a bunch of links it is reasonable to assume....
Oh, I am aware. Heck, the "quote a question and then give a bunch of links" isn't even required, since what "is reasonable" varies widely depending on the topic.
> ...that is the way they are answering the question...
But they didn't even try to answer the question that was asked. This is the beauty of just posting links: no claim of them being an answer is made, readers can assume for themselves that they have answers the question, and confirmed the meme.
> ...in short they believe that the links they provide are a good faith explanation of 'what degree "they" "influence" an election in our "reality"'.
People are welcome to believe whatever they like, but I am under the impression that what is being discussed here is at least an attempt at the truth. Could I be mistaken?
I will ask you point blank: do you care what the truth of the matter is here?
>You could also get some percentages with a random number generator. Are you asserting that reasonably accurate quantitative (percentage) truth can be derived from these articles? If so, I'd like to see you explain how, and also how you would determine your theory is correct in fact.
Ok there's a lot to unpack in what you said here but let's try .-
>Are you asserting that reasonably accurate quantitative (percentage) truth can be derived from these articles?
why would I? I didn't post the articles but anyway I think we are once again back to something you seem to like to do which is to demand quite a lot more work from other people than you seem to be willing to put in.
You made a short one paragraph statement that there was no proof of these things, the post then followed with a lot of links that I suppose the person who made the post had to gather in some way (perhaps they had it in their bookmarks tagged "election-manipulation" or something so it was relatively easy.
Then you claimed that was not good enough and that they didn't answer the question and then demanded they post things "In percentage terms please, and please explicitly acknowledge opinions as such." which is a lot more work than you put into that demand.
Then when you say
>If so, I'd like to see you explain how
Again it seems that you want me to read through all of that posters stories, provide you with statistics pulled from the stories (which will involve doing the actual work of putting the numbers in various stories into a stastical format) and an argument as to how they apply.
That is quite a lot of work to answer someone who posted a couple sentences, my advice is, if you want to disprove this person's assertions - do your own work.
>and also how you would determine your theory is correct in fact.
Well I probably wouldn't have any particular way of determining if my theory is correct because it would take time and time is a limited thing therefore you are probably required in life to accept some information you take in as true until such a time as you need to confirm, I don't have much to do with the Russia, or American Elections or any of that so I would probably just take in things that seemed reasonably argued as being truthful.
This means I might accept as true things that are false, and vice versa, because time is limited and I cannot fact check everything.
That said one way I might check if my theory was true is if people I disagreed with on the internet provided rebuttal arguments in the way of other links with numbers in them, because that would seem to me to be a good way of making an argument.
>People are welcome to believe whatever they like, but I am under the impression that what is being discussed here is at least an attempt at the truth.
Frankly it does not seem to me that you are making an attempt at telling the truth, it seems you are making an attempt at saying that something is false without providing any data as to why it is false. Maybe because gathering data is time intensive.
>do you care what the truth of the matter is here
Not really, I care that argumentation is clear enough that people can perhaps approximate the truth of whatever is being discussed through following that argumentation.
Please be aware that a lot of stories about bots on social media are themselves misinformation. If you trace them back to the sources you arrive at academic papers that are just intellectually fraudulent in various ways, like they misrepresent their data or they are identifying real westerners as Russian bots. I did some deep dives on this topic back in 2017-2018 or so:
So it's very important to be careful when making claims about "influencing elections" because there's such a long history of false claims from western sources, amplified by western media, of which the outlets you've cited are prime offenders unfortunately.
As this subject is interesting to me, I decided to read one of your articles to understand where you are coming from. And I came across this part[0]:
> The cited evidence is two men who joined the Lithuanian Rifleman’s Union, an organisation with a stable membership of around 10,000 people (0.3% of the population). ... The article presents no data or other evidence to suggest behavioural changes in the Lithuanian population: the anecdote of two people is generalised to the entire country.
So I checked the article and it says:
> "We are growing dramatically in numbers. Three years ago we had 50 people in Vilnius - now we have 3,000."
and then:
> Some 4,000 troops are being shipped out to the region - with 1,000 German soldiers allocated to Lithuania.
So while I believe your engagement in bot detection is genuine and you can offer some useful insights into how some bot detection platform are broken, as for the field reporting, forgive me but I still trust more traditional journalists than you, in spite of living on this Earth long enough to realize everything is imperfect.
The part about the troops isn't relevant given that it's a story about a civilian militia and troops is a reference to a professional army. Still, it looks like you're right about the recruiting. This was written seven years ago, guess I didn't spot the quote about their growth. I'll remove that story and add a note to the bottom. There will still be 15 examples which is plenty to make the point, especially as that's the least important example.
> I still trust more traditional journalists than you
By all means, trust who you want! But bear in mind you could easily check these examples because there are links and sources for everything, something journalists often don't provide. For example, the Reuters story in the parent post about "Peace Data" doesn't seem to have any links at all.
I want to believe you but now I have to trust your claims over what the NYT, etc are publishing. In order to do that I not only need to follow your sources but the original sources that NYT etc define to ensure they are the same. I also have to make my own conclusions about the data to see if I agree with anyone’s analysis. Frankly it’s too much to do when all I want is a few dopamine hits in the middle of the night before I can fall asleep again.
> In order to do that I not only need to follow your sources but the original sources that NYT etc define to ensure they are the same.
To put an even finer point in it, they're asking us to believe their blog posts over numerous large media outlets, and even above (presumably) peer reviewed research.
Anyone capable of understanding the issue of misinformation should see why that's a terrible idea, and why we shouldn't be encouraging that kind of behavior.
Right. The other thing is that the parent claims to be an insider which is great- I love that perspective. However, as outsiders maybe the researchers did the best they could with what data was available to them. After reading the blog article it seems like the main thing the researchers may have had wrong is the margin of error. I don’t have a huge problem with that since science is supposed to get more accurate over time as refinements come in.
Perhaps it would have been better if the blog writer had approached the researchers and said, “Hey I read your paper. I have insider knowledge and I’d like to help you refine your model.”
Hi Mike, so I'm watching what is happening in the misinformation are right now and while I don't disagree some people might have gotten some details wrong, somebody - or at least quite a lot of somebodys - is parroting the same statements that come from Kremlin on a mass scale. Some of these statements are completely untrue, many contains bits of real information mixed with false statements. I don't know who is behind these and I will never have a proof, but ignoring this massive phenomena will get us nowhere.
Specifically, what gets me worried is not the subset that is related to making Westerners averse to supporting Ukraine but many subtle and not so subtle attempts at sowing discord using the divisions already present in our society. This really works extremely well and we're super-weak when faced with these.
Does it bother you if it is domestic media and politicians sowing discord?
How do you figure all of this stuff out anyways, do you have a massive spreadsheet or model of some kind? But then, those can inventory stories, but they typically don't help much with whether the stories are true, assuming that matters.
Yes, domestic media and politicians sowing discord bothers me as much, also because it is done in cold blood - for money and power, respectively.
I only focus on fragments of reality I'm familiar with. E.g. I have a few friends in Russia and Ukraine and it's interesting to compare what's reported in the media with first-hand accounts (which can also be biased BTW). As for the rest, I'm humble enough to admit I don't know and may never know the complete truth.
The Russian claim that Ukrainians are Nazis. This claim is based off of members of right wing paramilitary organizations that fight Russian occupation. Of course if such a broad generalization could be made based on a few people then would also claim that Russia are themselves Nazis based off of Prigoshin’s SS tattoos.
One can find numerous examples of right leaning Americans repeating the claim that Ukrainians are Nazis. These same people are intellectually immune to applying the same reasoning when confronted by similar evidence of Russian Nazism.
It’s obvious that there are misinformation campaigns coming from Russia and China directed at Americans and others. That people suddenly no longer believe in the efficacy of the polio vaccine and other nonsense is evidence of the power of disinformation campaigns. Since they are effective, easy to do, and cheap it’s clear that all major state actors will engage in such things.
>Simple example: How do politically motivated troll armies influence an election in a reality where people don't use social media as their primary news source?
By having one super rich troll run for president and buy hour long slots of evening broadcasting. See Ross Perot.
I mean there's a world of difference between "Perhaps millions of people, given motivation and a bit of luck, can pull off a thing" and "A super rich guy can pull off a thing"
This is the whole reason people talk about middle-class modern humans of various recent eras "living like kings" in comparison to humans of less recent eras. The thing (Maybe that's "having people from across the world entertain you in your living room", maybe it's "Eating elaborately prepared deserts made out of refined sugar") was possible for some before, but not nearly as many. This is the kind of thing technology tends to change most often. Not best-case capability, but access and scale
Like some of those things are mass-produced superstimulus facsimiles and not good in the same way they were for the ultrawealthy of old, but even those genuinely are changes in how the world effectively works
> Each time I hear (paraphrased): "It is different this time.", I try to imagine the world 25 years ago. Then think: Is this really that different? Most of time, I think, "No, it isn't that different."
Just be careful not to base your conclusion on anecdata. There are legitimate differences amongst generations in addition to the many commonly held misconceptions, e.g. Millennials are poorer than Boomers were at the same age.
I'd you're interested, check out Jean Twenge's book "Generations" for a look at the differences backed by the best available data.
It's definitely a very similar method but fundamentally different in that the 'Distilling step-by-step' approach is a multi-task model.
As I understand it, rather that training the smaller model to produce the CoT/rationale as a part of the (decoded) response, it actually has two output layers. One output is for the label and the other is for the rationale. The other layers are shared, which is how/why the model is able to have an improved "understanding" of which nuances matter in the labeling task.
I suggest integrating this with every single help desk software you can. They all provide integrations. We were looking for something like this earlier in the year as it pertains to customer support and using SMS as a channel.
If help desk isn't the answer, then maybe another type of platform. Generally, I think you need to ride the coat tails of larger platforms.
(Note: I do realize this is developer focused today, but it didn't necessarily need to be.)
I strongly disagree with the sibling comment. Your website _does_ matter. At least in so far as it represents how you are positioning yourself to your clients.
Your website is all about what kind of software you can build. Your clients don't care about that. They want to know what kind of business value you can deliver. What kind of results can you achieve?
As a business owner, I don't want to see example websites, I want to see a case study about how you increased revenue by 20% with a new user onboarding process. I want to see how you saved 100 hrs per month in staff time (i.e. expenses) with the new set of automation features you developed.
If you don't have this data from past projects, then start collecting it. Start framing all of your conversations this way from now on. If a client is working toward a business objective, help them achieve it and make sure you put mechanisms in place to measure success and assign a real dollar value to what you helped them achieve.
Don't misunderstand me here -- contacts to a employee are not bad per se, but in my experience, those contacts usually care about their problems, but do not have the budget to have the work done properly.
OTOH, if you have a contact that is higher up in the chain, those tend to have a broader view and you can discuss prices not in terms of effort on your side but in terms of business value for them.
