I used AI to analyze presidential speeches... I was surprised
Вставка
- Опубліковано 16 вер 2024
- In this video we're going to be looking at presidential inaugural speeches to see which are more positive and negative. We can plot these over time, contrast political parties, analyze re-election speeches, and we can even ask the AI to tell us what it thinks are the most positive and negative parts of both Trump and Biden's inaugural speeches.
Thanks to my patrons for their support! You can get your name in the credits too:
/ datatime
Raw Jenky Data:
docs.google.co...
References:
platform.opena...
www.presidency...
pypi.org/proje...
en.wikipedia.o...
www.clarkart.e...)
en.wikipedia.o...
slate.com/news...
Thanks to Storyblocks
Special Music Thanks:
AM 1984
Udio, AI generated music
Is NOT what political party speaks more negative that is all about, but rather who speaks the Truth and honesty ,without deception or malevolence ❤❤
Trump 2024
@@rudeR6I agree, Trump is the most malevolent liar this country has and I'm shocked he's still allowed to run
@@MawdyDev what did he lie about?
@@rudeR6 he said he would fix the things he mentioned in the speech mentioned in this video, then proceeded to make them worse.
He blamed problems that he caused on other people, such as rising gas prices which he caused by making a bad deal with Saudi Arabia and then blamed it on Biden, he worsened the COVID response by telling people "it's overplayed" and telling people to use pseudoscientific "cures" instead of an actual vaccine, then, again, blamed Biden for the aftermath.
He defamed legal workers, and despite his lawyer admitting outright in court that the statements were defamatory, continued to express that they were true outside of the courtroom.
In the 90s he claimed on a talk show that "you could tell Christians anything and they'd believe you as long as they believed you were also faithful." That one has me the most angry, since my dad who's Christian fell for his lies.
And I don't know what you would call Logan-Pauling a military funeral for political gain, but it's definitely not a good thing.
He claimed he wanted to fix education, then started slandering underpaid and overworked teachers instead of lightening their already excessive workload.
He wants to ban abortion but has no interest in keeping children or their mothers alive after being born, it's almost like he wants them to finish cooking so whatever demon he worships can have a better meal.
The one thing stopping major illnesses from being cured despite having the technology to do so is the profit motive to keep people sick, and Trump plans to block all attempts to remove that malicious profit motive.
My stepmom has breast cancer and my dad still worships the idol that plans to keep her miserable.
Even after he called veterans fools for not skipping the draft like he did, even after he desecrated a military memorial, even after everything Trump has done to actively harm veterans while claiming on Fox "News" to support them, my dad still follows him.
Trump brainwashed my father and I can't let that go. No amount of logic is getting through to him. I should've child locked Fox "News" on his TV the last time I visited.
@@rudeR6 The list got removed, I don't know whether by UA-cam's automod or by the channel owners, but I won't bother typing it all again.
I'll leave a brief footnote:
- Admitted in court that he defamed jurors and poll workers but continued to deny it outside of court
- Made a bad deal with Saudi Arabia resulting in skyrocketing gas prices, which he then blamed on Biden
- Pushed pseudoscience and told people not to quarantine, which drastically worsened COVID in the following months once Biden took office
- Said in the speech in this video that he would improve education, but over the last 8 years has done nothing but make teaching such a financially unfeasible career that we now have a nationwide teacher shortage
- Logan Pauled a military funeral then claimed that the worker who tried to stop him was mentally ill
- Claimed that he supported veterans but called them fools for not skipping the draft like he did
I'm going to copy this one to my clipboard this time, since I don't want to re-type it every time it's removed. People who haven't been told this need to know, even if it's more profitable to keep people misinformed.
It’s not so much measuring their reading levels, but their speechwriters’ writing levels.
I have seen a video about the AI analysis of the 2024 trump-biden debate. That was something surprising.
More likely the reading level the speech-writers stategically decided towrite in
When you make a speech, you're first job is to know your audience. The speechwriter is tasked with this expectation. Clearly, Biden expected a dumber audience than Trump expected. Of course he did: he's a Democrat.
Variations in speeches could be affected by whether or not they wrote their own speech or had help. Also generally round up when fraction is over half. So Biden would be sixth grade level. And that could be by design to reach greatest number of folk.
@@janiceb8809 Not saying much. Education as poor as it is in America. In the 1960s, education was far better than today -- America was actually #1. But today, America is #24 in reading and science, and #36 in math. But half a century later, I realize that education in the 1960s was awful compared to that in the late 1800s. And that was likely poorer than the early 1800s before Progressivism polluted it.
The biggest regret I have in graduating from the 3rd highest scholastically rated school district in the country, and in the top 10% of my class, is that Civics was virtually empty of value -- with virtually nothing of the philosophy behind the Constitution. And my first 2 years of Spanish class had horrible teachers; only the 3rd year had one of my only 2 excellent teachers.
😎♥✝🇺🇸💯
Using a speech to determine a President’s level of speaking does not determine the level of a President’s speech level. They all have speech writers who write to convey a message that the public understands.
The reading levels of Americans taught in public schools has been falling ever since the Dept. of Education was created. Speechwriters craft their speeches to appeal to the sensibilities and education level of their target audiences. Compare the speeches at public rallies to their speeches at exclusive high end fundraisers.
@@luddityEXACTLY! They have really “dumbed down” the citizens! By design. CORE was created, according to the “creators”, because they just want similar “workers”! (Basically, ones who won’t think on their own, and do as they are told!
@@luddity Well, I suppose you think that without the Department of Education, things will improve? Yeh, right! There's only ONE person who said they love the uneducated. and who wants to do away with the ministry of education. the one who moans and groans and spews random words but thinks he has the best words.
It sounds like it very much does.
After all, they're just analyzing the level of the inaugural speech, not the president's overall intelligence level.
Exactly I think this is interesting and I would love to see it done with maybe the last 4 or 5 presidents when they were just taking questions from media that would be much more interesting than a speech given like you said that is not from them it's their writers that do
I worked for a Medicare contractor for 12 years. When I first started, we had to make sure any letters we sent out were ranked at 6th grade reading level. By the time I left the job, it had dropped down to 3rd grade.
I have an Etsy shop and when I add new listings, the instruction is always that my language is too advanced and I should use simpler words for the mostly US audience. I find that difficult as I'm English so want to use the best descriptive words I can!
That’s very sad. 😢
So, are you saying a persons value is tied to their reading level?
@@thewatchdogs1109 That's a total misreading. So why are you introducing a straw man?
The grade level decline is not accidental.
When Washington spoke he was speaking to the most educated people in the country who just all happened to be in the room with him, since the invention of Radio TV and internet, these speeches were to a wider audience of far less educated people who populate the whole country so shorter words consistent with average day to day conversational speak were needed for the masses to know and pat attention to what is being said.
It's also, as noted in the video, an artifact of using a modern analysis on archaic speeches.
Well said!
They did also speak outdoors in public on occasion. Gettysburg Address for example.
This is fair. I’d imagine that we would see a similar trend on them speaking off the cuff. We used to elect the best and brightest and now… 🫠
YES. You can't talk over people, you need to speak their words. Proving you've read a dictionary isn't always the point.
An ai version where it compares presidents lies vs truths would be the ultimate analysis.
At that point, I'd question whether the AI is qualified to distinguish the truth from lies. There's no mathmatical equation to determine how true a speech is, unlike the grade level of a speech
@@joncents2000you just need to ask it in a fair manner. Such as is this verifiable? (Do not stop there...) According to experts, is it honest? Is it fair? What does this mean for me, my family and the average person? Is there a story behind this that I missed?
What else do I need to know about it? Why are other people calling this a lie? Are they wrong?
I never seek confirmation of my beliefs because my opinions can't be relevant unless I have all of the story. We used to get more of the story than we do now and I only follow accounts that use the same methods.
@@rachaelkp unfortunately this still leaves a lot of room for the water to be muddied. The biggest problem with defining "truth" is that experts are funded to study things, there for there is always implicit bias, who was funded, who funded them, why they wanted the study done, how the study was done, it's never just as simple as "there is a study done by experts so it's true". Not to mention government and corporate intervention and censorship that intentionally aims to hide or destroy information that goes against their agenda. It's easy for history to ignore things like that the american heart association was funded in order to create studies 'proving' that vegetable oil was more healthy, and running advertising campaigns based on the AHA findings to claim that competing oils and fats were dangerous, which today we know is not true. Similar things have happened with wheat, climate change, covid 19, red meat, chemical waste dumps, basically anything and everything there is always situations where "truth" given enough time can turn out to be corrupt organizations intentionally creating data sets that are designed to prove whatever "truth" they need to benefit them.
Especially when it comes to training AI on what is truth, the information they are fed is what defines the truth, and that information will almost certainly be tainted by censorship of information. Sadly in most situations truth is rarely objective, and almost never a simple is or is not true, there's often much more nuance that needs to be considered.
AI cannt tell truth from lies
It would be subject to the training data that tells it what truth is
The best communicators are the ones who can most effectively explain complex ideas in a manner that lay men can easily comprehend.
I agree.
Like Vivek
Fifty years ago we were taught that most Americans read at a sixth grade level. So if you're writing, mAke it more understandable.
@@bren.r And Reagan.
Yeah, what he said.😊
George Washington’s speech was written by James Madison which explains the high grade level.
Dude was a genius.
AI cannot deal with sarcasm.
its getting better but nope not even a little has trouble with satire as well
mine can. detects nonsence. dont go chaceing waterfalls
@@grummbunger❤ my AI is common sense and Christian values.
@@terriodom4266what are you yapping about ☠️
@@duckified. 🙏🙏🙏 for your soul🙏🙏🙏 GOD is over all!!!
This video should have said, "Speeches are just words, and words can make sentences and could be true or lies. Always remember: Actions speak louder than words."
Words are very important! Hate crimes and political violence increased 4:18 when Trump started campaigning for the presidency in 2015. That trend continued in 2016 ... Its climax: January 6. Words can hurt, words fire up emotions and lead to acts
@@gabyschilz4484 Good to remember, the idea given in words lead to action generated by the words. Here's to encouraging words of peace and reconciliation by all for all.
@@gabyschilz4484 Yes, I remember all of the lawlessness and violence from Antifa and BLM. Scary times that were brought about by inflammatory misinformation and hateful rhetoric from the main stream media, as well as those in politics. Time will tell if the media will change their manipulative reporting which has been intentionally trying to create fear and hate in one group of Americans toward another group of Americans. We are all brothers and sisters. We shouldn't let anyone divide us. Don't believe the lies. It's important to look at different sources in order to get a fuller and more accurate picture of reality.
@@gabyschilz4484 So therefore, if we want world peace we just need to police what everyone is saying. Or even better, what they are thinking, and very soon we can do that more effectively than at any other point in history.
Utopia is in reach, the law just needs to catch up to technology.
@@Reddotzebrayou should have added "warning: this post contains irony" just in case. Wouldn't want to cause offense.
A.I programmers biases have been found in algorithmic coding of A.I data to subversively present it as intellectually unbiased. Your engine will reflect it's engineering but reveal the engineers limits.
100%
Exactly. As it associated "democracy" as a "positive", when it is essentially a glorified riot; mindless mob rule with extra steps.
🎯🎯🎯
Not necessarily but mostly true. When all the database is your own and if the AI is basic it will work. There are some open source AI tools that could do this
Someone else who's clueless about how AI is coded.
This is the first video I watched from your channel, and I have to say I do appreciate the non bias approach you have taken to political sides.
No check of lies though
I guess ‘I beat Medicare’ would be ranked positively.
We finally beat Big Pharma and gave Medicare bargaining power.
He just couldn't get all the words out over Chump's yelling.
Badacafcare, nexnelrecent, truininashabadapressure
Also boost his word length for his Kincaid score
😂
But negative when Trump responded and told the truth that yeah, "Biden beat it to death". 😂😂😂
Lol
Yea guys, feed it unscripted material and not speeches written by others and everything changes drastically. Do this cause we all wanna see it!
🎯
You noticed that too. That speech by Trump is not consistent with his usual public speaking style.
@@mightymystery9204 That was the first thing that came to my mind!!
then we cound be who saw what could be unburdened by what had been.... I'd loveto see those numbers
Trump comes in at 0.4
Truth can sometimes be seen as negative. That’s the truth
Correct. If big government is considered a positive thing and considered the truth; anything advocating freedom, liberty, & independence could be seen as a negative thing.
Lots of people who claim they want freedom, turn around and say they want to ban 50 things.
@@Seth9809
When they tell you who they are, believe them.
Tyrants wear many hats/faces/etcetera. They come from all walks of life.
Don't be so negative.
@anthonylosego Why so down on other people's negativity? Additive negativity isn't the same as multiplicative negativity. It still results in a negative. Just sayin'.
It's shocking to hear the difference between Trump's Inaugural Speech and the way he talks now.
As aligned with the lack of progress as measured by outcomes from his time in office
He didn't write a word of it, and had to stick to the script. As soon as the teleprompter was turned off, the real trump came out and refuses to leave.
@@margaretsomerville2510 He did accomplish alot while in office. I hope he gets four more years. Trump 24!!!
Why do single older women hate Trump? The rest of the Country loves him. What is it about unmarried menopause women that finds Trump so bad?
@@DL-wm5quobviously your talking about Biden, or Harris. That's why they don't do interviews. Trump will go anywhere and talk to anyone, even has his debates on CNN, and ABC the most hostile news outlets toward him. Have you ever seen a libtard on FOX?
If a president delivers a speech written at a 5th grade level and butchers it shouldn't they be deducted a grade level?
Or salary?
Biden isn’t being fed.
5th-grade level was far too generous for Biden. 90% of the utterances he's made in the last few years have been incoherent nonsense. Gaff's, stuttering, slurring, lapses, erratic modulation control, and on and on... We've all been forced to ignore the fact we're witnessing the real-time collapse of a brain suffering from dementia, or perhaps Parkinson's, or some other neurologic condition. This was all obvious well before the election. It's been borderline elder abuse that they've even allowed him to speak at all. It's sad more than anything else, a true reflection of the decay in politics and in so many of our institutions.
Lol 🤣
@@realityobservationalist7290
Well said
The reading level of speeches may depend more on what the speaker expects from his audience than on the speaker's ability. (Most speeches are written by professionals with an insane reading level anyway!)
Early presidents, for example, addressed their colleagues who were present at the inauguration, later presidents addressed the entire population reached through the mass media...
To summarise, I would argue that the trend in the speeches is towards increasing POPULISM.
That is quite a risky conclusion/summary. Gather some more data and make a better case for that claim, chop chop!
@@justsayin...1158that comment read more as a hypothesis than a conclusion and or summary to me at least
Among other reasons, it realy seems so
But during past times, if a politician would have spoken with a low reading level
they would have lost credit a lot, even when adressing the public.
Also, languages have tended to simplify over time, especialy the english language.
The media also allows for easy to grasp concepts to spread much faster,
which would naturaly lead to oversimplifications by politicians.
But facts* are also much easier to look up, likely leading to the more widespread development of "alternative" sources and facts*, adressed as such.
Generaly, the more people felt like they need to control the public and convince them that the established is wrong, they turn to populism.
But generalising, even if sadly pretty much needed because more will hardly reach the public, is always flawed in complex concepts like sociaty and politics
I tried to articulate a similar comment but I missed your crucial detail of writing for your audience using that phrase talking about television.
They don’t write their own speeches
That WH Harrison's was the longest is amusing, as he was the first one to die in office, died because he caught a horrible cold by making his speech on a cold, wet day in the rain, for *two hours*, and refused to wear an overcoat or a hat.
There are questions about this actually being the cause of his problems, but it was likely a contributor.
Colds are caused by viruses, most commonly rhinoviruses, which spread through person-to-person contact or by touching contaminated surfaces. President William Henry Harrison's death is often cited as an example of this misconception. He gave a long inauguration speech in cold, wet weather and died a month later, but historians now believe he likely died from pneumonia caused by bacterial or viral infections, not from exposure to the cold alone.
Being in the cold does not give you a cold. Coming into contact with the rhinovirus does.
Yeah, too bad the common 'cold' is caused by a virus and not actual cold or rain. He would have gotten sick and died from the virus even if he wore a hat and coat. Viruses and bacteria cause diseases, not not wearing a hat or coat. That's just an outdated wive's tale.
@dvig3261 Yeah that's what they say, but I think it's not true. Being cold can lower your immunity. Personally, when I get chilled, I often get a cold. And if there's a cold draft on my head for more than a few minutes, I always get sick.
@@dvig3261 It doesn't give you a cold but it can make it harder for your body to fight it.
"In Springfield, they’re eating the dogs, the people that came in. They’re eating the cats. They’re eating, they’re eating the pets of the people that live there."
F-K score of 0.6
"Ukraine is a country in Europe. It exists next to another country called Russia. Russia is a bigger country. Russia is a powerful country. Russia decided to invade a smaller country called Ukraine. So, basically, that's wrong, and it goes against everything that we stand for."
Kamala, 1st grade
@@Uiopuiop123 4/10grade above above trump and something that is both based in reality and correct. I'll take it!!
@buckcherry2564 😂you don't make sense
@@RLWizard You not understanding is not on me.
@@Uiopuiop123 That's beautiful.
It's the speech WRITERS are the ones you are assessing, not the presidents.
but the presidents approve them, and actually get up there and say the words... Thereby assuming ownership of the speech.
@@genxer74 LOL. So, it affects the stupid implications that so many ignoramuses are tossing out. It's not useful analysis for anything, really. This kind of garbage is useful only to Hollywood screenwriters in character development. Fictional characters.
Trump doesn't use a tele prompter so they are his speeches.
@@Mclvn21091 sure buddy, keep telling yourself that
he got someone to write his own book. he gets someone to write his speeches.
@@Mclvn21091 "I pay all this money to teleprompter people, and I’d say 20% of the time, they don’t work,” Trump said, adding he would not pay the vendor who provided the prompters. “It’s a mess.
June 9, 2024
That's a very odd thing to lie about and very easy to check.
I have HUGE respect. You analyzed the data of Trump and Biden but didn’t say one side was awful. In the world we live in, a simple thing of giving facts without all the extra stuff is rather refreshing.
Thanks!
@@DataTime27 you guys basically did what the Media is supposed to do. Give the facts without injecting opinion. And you did it masterfully!
Thank you so much!
@@DataTime27yes you did. Hence the negative one which was a lie. Bidens speech with the red backdrop can't be beat with the negativity level.
@@MrMcGribbleno they didn't. They gave their OPINIONS AND LIED ON THE NEGATIVE LEVEL. Yóu can't be at Bidens negative level speech with the red backdrop.
Can we also do where we only analyze presidents freestyle by not including their writers speaches.
This would be great. I wanna see what they do in unscripted interviews and such.
Yes, how would AI like speeches that include sharks, and Hannibal L? And would AI think about all the slurring? Lol
It wouldn’t work. For example, Biden don’t do phrases without a teleprompter
@@houlej19 there are no teleprompter on debates
@@faikerdogan2802 that’s why you couldn’t say a full phrase
The reason the speeches having changed in sentiment over time in general is because the State of the Union and inaugural speeches were as much a news program, as a chance to hear from the President.
The most surprising thing about this video is the comment section is calm and collected for a political theme.
Impressive.
But I was also suprised what the data revealed.
I am surprised that anybody claims to be surprised.
Fact 1) Both sides are somewhat the same
Fact 2) Trump spoke at a higher level than Biden. This is based on both the cognitive ability of the speakers as well as the adressees of the speeches (smart workers vs hoodlums and welfare queens).
Fact 3) The inaugural speeches were more positive. Uhhh, they are in office now and try to motivate and energize the listeners, d'oh.
Nothing in this is really surprising.
True. A pretty neutral video
No way! Kamala is a communist and Trump is literally Hitler!!
I think people interested in data are a little more level-headed.
Challenge accepted…BOO!! We are the “correct” ones bc look at the color of our neckties.
Interesting work, but important to point out that speech writers deliberately target a lower grade level, sadly because that represents the average voter.
The problem with using AI evaluation is that AI results are skewed to reflect the programmers' biases.
Most AI programmers are heavily biased against Trump, yet this AI slightly favors Trump (if you believe reading is a good thing). Something to think about.
@@awesomeferret they only fed the ai with the inaugural speech. that wasnt trump's typical rally speech.
@@cerebrumexcrement I'm very curious to hear why you said that as if you thought I wouldn't already know that. What was your thought process in assuming that I wouldn't already be aware of that?
You have no clue how AI works. Programmers can't instill their biases into AI because AI isn't coded like you think it is. These models learn based on thousands or millions of gigabytes of data of all kinds and what they "conclude" from the given data isn't predictable. Why do Americans impersonate the Dunning Kruger effect with such entitlement?
@@awesomeferret😂😂😂
your intro has some real 90s science class movie day vibes
Of course that computer had to be informed on what was positive or negative.
:D
Yes, it that’s been done by hundreds of people and tens of thousands of speech snippets over several years. Sentiment analysis is pretty well locked down now.
Google Large Language Learning Model....
The language model is trained in English compression.
It's still not perfect. Janelle Shane wrote of an AI that was given hundreds of restaurant reviews and asked to rate them as positive or negative. For some reason, it decided that all reviews of Mexican restaurants were negative. Analysis showed that the Large Language Model considered "Mexican" a negative word, because it was often found close to the word "illegal."
This analysis only reflects what the speechwriters wrote and the speakers approved.
It also reflects the bias of the Programmers & the Scientists behind the Word studies.
It’s understandable that modern speeches are targeted at a lower reading level.
It’s not that the leaders or speach writers are now dumber, it’s become recognised that you need to speak to the common denominator of your audience or risk losing them. Relatability has come to matter more. So speaking using simple, common words in short more memorable sentences has become a best practice.
Of course, this is different to incoherence. A good speach is both coherent and relatable. To its audience.
Indeed. Another important factor is that the population gets older and older. With that comes both difficulty hearing and cognitive impairment. Also, in this interconnected global world, an American president of today not only speaks to people whose first language is English, but with people who have English as their second, third or fourth language. As a disabled person, I feel very strongly about inclusive and accessible language. Just because some of us need simpler language doesn't mean we're stupid.
The ultimate goal of a good speech is connection with the audience. Otherwise, a speaker speaks to hear himself speaking and becomes banal to his audience.
The real culprit in today’s society is the extremely poor quality of public education. When a majority of voters are ignorant of the public policies that affect their daily lives and are swayed by eight second sound bites, we get the government that we deserve. Until and unless the education system is dismantled and rebuilt, the United States as a society will continue to decline.
You guys are like bringing back the 1980s how you guys open up honestly I dig it
"Brevity is the soul of wit." - Shakespeare
I would not trust AI to evaluate anything. Given what we've seen about the clunkiness of AI, I would not trade in my own reason, intuition, and analysis to machines that have none of those things.
I'd trust AI over main stream media or reddit comments... which is where most people get their info from
@@j44m33 Have you used AI much? They're good at creative writing but terrible at math.
@@valleyshrew daily, id still trust AI over MSM or reddit. or you have better alternative?
Yeah, you're hanging on to denial. The best publicly available AI chatbots are capable of far better linguistic analysis than the average person, right now. And very soon they will surpass genius level. No, I do not give a crap about the occasional error. Humans make FAR MORE errors, and no human is judged on their work, seconds after their first rough draft. We revise and rewrite before submitting. If you allow Claude or ChatGPT 4o to do the same, there is nothing clunky about them and it can all be done in minutes or seconds instead of hours that a human would take to do the same work. I know pooping on AI is very popular right now, but try thinking clearly and learn more about the topic, and you will know better.
As someone who worked in the computer field in the early days, the same adage about data holds true...garbage in, garbage out.
Ok. This is excellent content. Unique, relevant. No political spin (mostly). Well scripted. Great sound. Smooth transitions. Just enough humor, yet stayed on track with the thought process. Really good, fellas. I think I might actually share this video. Not because I will make a difference. I'm a nobody. But it was just that enjoyable and I haven't thought that about a youtube video in ages
Awesome! Thank you so much!
Gosh - both of them have aged SO much.
This was a truly fascinating video. I wonder what would change if we compared Trump and Biden speeches to their regular, unprepared, unprompted speaking. It'd be cool to somehow get the captioning of them perhaps answering interview questions or simply talking. Then comparing to their written speeches. I imagine the difference would be interesting
"Trump in Biden speeches compared to their regular, unprepared, unprompted speaking"
There is a video on YT about the 2024 trump-biden debate, AI analysis.
@@williamforsyth6667 do you have a link? I'm struggling to identify the video
@@GabrielsLogic Sorry, it was weeks ago, I dont remember the channel, and could not find it with a simple search.
umprompted speech by Biden: ?..................?....................?..............
@@williamforsyth6667 it's all good, it must be floating around the Internet somewhere lol!
One thing I realized about Trump's way of speaking when comparing older videos of him, do not lose meaning when translated to non English speakers, or, easy to understand for people who partially speak English.
a skill i would expect for some one that makes billion dollar deals in other country's
@@Loan--WolfOr tons of people in a country which LOVES importing people from other countries (The United States of America)
@@gooberaureliusmaximusiii6638. Good point but I definitely don’t LOVE importing illegal aliens into any country, let alone the US.
Interesting find
@@gooberaureliusmaximusiii6638 The US was built by people from other countries.
William Henry Harrison's speech was almost as long as his presidency.
And probably the cause of his tenure's duration.
Having done my master thesis on discourse analysis, I'm simply blown away by the possibilities you just showed!
The contrast between his speech at 11:10 ish and the mumbling, random noises he makes now is ASTONISHING!
If nothing else, it truely shows how severely his dementia, and other health issues, have affected him!
The D word is very applicable to recent politics but if you don’t get that it only applies to the blue guy, you better get a cognitive check yourself.
@@stevethompson210
If you're trying to say what I think you are, you may not be suffering from dementia, but you certainly have a dissociative disorder.
I had forgotten that Trump once spoke with some eloquence. Now it's all complaints, nitpicking, and rambling off topic using repeated superlatives.
@@stevethompson210To be clear, he was referring to Trump.
@@alanjonadenz-hamilton3727 eloquence is not a word I would ever attribute to Trump, past or present.
When the state of our country is negative... I would expect speeches to mirror that.
Its always negative. Even if it's not, it's always negative.
We have lower crime than under Trump and a much better economy, the best on the planet. Data supports that assertion.
Fascist dictators use negativity to create distrust in institutions as a pretext for their seizing of power. Remember, Ben Franklin advised we never give away our liberties for promises of safety, for we’d end with neither.
There are always positives and negatives. All negative is not realistic. It is a common trick of people who want to be authoritarian dictators to present the country as an entirely negative dystopia, and present themselves as the one person who can fix everything, and save everyone.
@@daniellamcgee4251 Straight out of the Marxist handbook.
Inaugural speeches would naturally be more positive as it is a beginning.
After spending last year working with dozens of AI models including Chat GPT, I have found that they are frequently wrong, hallucinate "facts," fail to follow instructions, and lack basic comprehension skills. IOW, this is a clever idea on its face, but I doubt the results are meaningful... unless you yourselves did most of the work in order to keep the AI on track and somewhat accurate.
As it’s a basic analysis of a few speeches, it’s less likely to show bias than when asking so to create something…
Apply the same analytics to the mainstream media and we could have an interesting video
We’re not talking bias, I’ve had AI models not sum a list of accounting numbers correctly.
It can’t add properly.
Okay I know that doesn’t seem like a big deal, the 0th instruction for every ARCHITECTURE level framework for computing is the add instruction.
It communicates the logic that instructs the computer to move 1 register down… literally every time the computer is turned on it is taught to add down its stack to retrieve instructions…
And the most complex version of our AI can’t do the first task of any computer… that’s worrying.
But idk doesn’t bother me I make a great living off computer science.
A.I. is the biggest scam of our times. Tech companies are now calling almost everything A.I. Just because an app is able to edit a picture doesn't make it AI. Photoshop has been doing that for decades.
@guyincognito1406 this. It's a scam.
I have to really commend how unbiased you guys come across as. I was expecting there to be some very obvious preferences slanting one way or the other, but it remained very centrist and neutral. Well done!
Thanks! That's what we should hope for
I find the correlation between positivity and reading levels really interesting. FDR was more positive with each speech but had a lower reading level while the Newer president showed the opposite correlation maybe there's something here or maybe I'm just grasping at straws
Yeah it's really intriguing
I think a part of the reason to blame is because usually more wiser words are more neutral, and this can be used well to describe yourself as less involved in the nation's bad times, or more involved in the good ones. Sorta like "Thanks to MY work the company survived"
@@Idoexist._. I think that's an interesting way to put it
Interesting, but grasping. It's easy to get lost in the weeds of correlation where zero causation exists. Take the climate hysteria, for instance. There's a real data nightmare, based more on *_feelings_* than on reality. And I used to believe the hysteria, until I dug into the *_data._*
*_Ice Age Madness_*
We live in an Ice Age and the "elite" are worried about global warming (Hint: They're *_not_* worried, with Obama buying sea level house in Martha's Vineyard; either that or he's stupid).
*_Garbage-in = Garbage-out_*
Big Lies feed the data engine -- like warming causes more violent storms and droughts, when Truth (data, evidence) tells us the opposite. When there is no correlation (coefficient close to zero), there is no causation between the two elements being examined, at least on that time scale. But even when there is strong, consistent correlation, this does *_not_* tell you the direction of causality.
*_Direction of Causality_*
And that BIG graph shown near the beginning of Al Gore's big fantasy film? It shows CO2 and temperature with a consistently strong, positive correlation right up until -- *_but NOT including_* -- the end, where CO2 skyrockets, but wimpy temperature is left far behind in the coldest major warm period (1,000-year cycle) of the Holocene interglacial. And the Holocene is *_not_* the warmest interglacial of the Pleistocene.
*_Fancy Theory Based on Accidental Correlation and Backwards Causality_*
The so-called "greenhouse effect" (GHE) is based on the absorption and emission spectra of certain gases which are, in the visible spectrum, transparent, but opaque in some infrared frequencies. In other words, the active ingredient in the GHE is light, which travels at 186,000 miles *_per Second!_* So, we should *_not_* have to wait decades or centuries for GHE to warm up the planet to the degree suggested by Gore's BIG, skyrocketing CO2 graph. Instead of the +17°C of estimated, expected warming, we've only had +0.8°C in the 20th century and the warming is *_slowing down!_* Nature had already scheduled warming for the 20th century, similar to the Major Warm Periods of the Medieval, Roman, and Minoan periods.
Gore's graph shows the only time scale with a consistently strong, positive correlation, but he had it backwards! Temperature drives CO2 into and out of the oceans as a dissolved gas. In fact, there is about 50x the CO2 in the oceans as we have in the atmosphere. So our CO2 could be as high as 20,000 ppm if the oceans were forced to release all of their carbon dioxide. If the GHE is real -- and it might be -- then it is *_too wimpy to register!_*
Truth be told, I fell for Gore's Big Fantasy. And only in 2015 did I wake up to the lies I had been sold by a slick-talking politician. My #1 weather bestseller, *_Climate Basics,_* tells part of the story. My other weather bestsellers, *_Thermophobia,_* and *_Red Line - Carbon Dioxide,_* complete the picture.
😎♥✝🇺🇸💯
I'd expect some correlations there, since "reading levels" are really just a measure of how long your sentences and words are. This will inevitably have some correlations with the sentiment expressed in such sentences.
We may be inclined to appropriate extravagant verbiage in matters of incremental complexity, so that we might appropriately task our conversation partners with the nuances and details of our situation, bearing in mind the multiplicities of factors surrounding the topic, enabling us to reach a sophisticated conclusion; just as we might employ verbose sentences in an effort to not enunciate duly the extent of our mismanagements and shortcomings, in which we employ overambitious nuance in an effort to obscure the truthful nature of our impending situation; just as, my comrades, it may be tempting to initiate grandiose eloquence in the pursuit of appearing smart.
Or we can use short sentences. They make for better one-liners. They are punchy. Easy to understand, so they're more suitable for regular folks, saying "I'm one of you!". They are clear, sharp, good to make a point. You feel it. So to really say how bad things are: keep it simple and stupid. And while flowery language can be inspirational, creating a crescendo of optimism to carry positive sentiment among the listeners-on, it nevertheless may not have the impact of a simple "Make. America. Great. Again!". Or even, "U S A!"
This analysis of inaugural speeches is focused on the most prepared, professionally written and rehearsed speeches - I see big differences to the routine extemporaneous campaign speeches particularly wrt the nonsensical sentences, the incomplete thoughts & statements, the self-interruption, the context changes mid-sentence, the mispronunciations, non-words and non-sequiturs and the bizarre insults and name calling. These contribute massively to perceived reading levels & general decorum too - but this algorithm might perhaps ignore them anyway?
i am positively suprirsed to have found a channel that actually makes no attempt to color the data they present with their own biases in either direction. Bravo!
Thanks!
Yeah. I agree. I was hovering over the subscribe button waiting for the obvious bias, but it never happened, so I slammed it.
The (unintentional) bosses occur from the input data. Reading levels mean nothing because the speeches that they present were (most likely) not written (in part or whole) by the President himself. There are a myriad of other data that could have faulty or misleading comparisons. There is no way in hell that Trump can have greater reading comprehension than any other President
Ok, the negative speeches correspond pretty well with wars. You have the war of 1812, the mexican war, the civil war, spanish American war, and WW1.
I believe that the speeches reflect the uncertainty around war
I appreciate the unbiased analysis!
Thanks!
Ha! Open AI has a liberal bias. So, sorry NOT an unbiased analysis!
Unbiased? You're analysing speech writers ability to satisfy what points the candidate wants stressed.
I think you shouldn't compare D Vs R in current time to before when the parties swapped sides.
@@LeifNelandDk When exactly did they swap sides? Senator Byrd was a democrat who got his start in the KKK. Same for the Kennedys and they're still democrats. Quit stating false information as fact.
100 years from now: AI analyzes how many words besides "like" a president knows.
I would be curious to see how my own daily speaking style is graded. This was honestly a very interesting video and I appreciate the straightforward nature of your presentation. Thanks gentlemen 😎👍🏼
Thanks!
The inaugural speeches are not a good example of a president's speeches. They are carefully crafted by highly rated writers. You need the regular speeches to find anything.
Trumps sounded like himself, Bidens sounded like what his owners wanted.
@@tommartinez62 That is not what Trump sounds like when he speaks “off the cuff”. Those are not his words. Those are the words of very talented speech writers. He held tightly to the script, looking from left to right, from Teleprompter to Teleprompter, and didn’t miss a beat. It was quite impressive.
ITS DATA TYIMEEEE
Ask ai to analyze the promises in the speeches verses what actually happened if they were elected.
Another one that would be fun is to see who wins most often negative speeches or negative speeches
I didn’t actually say that right.
I thought that as a politician Biden spoke at a higher grade reading level. But no. The guy is truly a fool far and beyond.
"81 million votes" my ass.
The president doesn’t write his speech. It does tell you a bit about the Party but it’s not the Nuts and Bolts.
Trump probably writes most of his own
The excerpts illustrate the limitations of Flesh-Kincaid Grade level for speeches.
Political speeches are *littered* with lists of parallel or near parallel clauses. (I blame Churchill.) Whether those clauses are *sentences* essentially determines the reading level.
In the quote at 12:04, the reading level goes from 8.3 to to 18.9 if you switch all the periods for semi-colons, and you’d be totally justified to do so.
OTOH 11:30 puts semi-colons between the items. It’d go from 17.7 to 9.1 if they’d treated them as distinct sentences, or as bulleted list.
Every single quote from trump & biden in the video has this pattern, and the punctuation chosen has a bigger effect on the reading level than anything actually in the speech.
(This post is grade 5.8 btw.)
Yeah I wonder if I should use a different score instead of Flesch Kincaid
Plus it doesn't properly qualify the complexity of the underlying ideas being communicated. I'd argue that a ratio between the complexity of a concept and the "understandability" of a phrase might be a much more significant measurement, if only it were not so subjective... maybe another idea for AI to grade?
@@DataTime27you should analyze how they problem solve, and whether their solutions work.
Communicators should always speak in the language of the audience. Scoring the language is useless.
Ive been thinking of making essentially a video essay about how politicians are getting more personal and also more polarizing, which is quite a similar topic to this.
This video was amazing though and you brought the point across in a nice and neutral tone. Good Job!
Thanks and good luck!
Impressive video that hits pretty nicely! I love the 90's kind of vibe it gives off from your production. Subscribed.
The worst analysis I ever see in my life. Talking about pandemic was positive, not negative.
I'm currently on my masters thesis, analyzing a novel problem. I am so lost in all the information. This video is so inspiring and urges me to dive into any foggy problem and get the most out of it. amazing content, got a subscriber here
be careful not to use unscientific metrics like this one. Longer sentences and words don't make for quality communication. As someone who's in the writing and sociology spheres, I can say without the shadow of a doubt that as a writers gets better at their job they tend to cut on the fat to get to the crux of things and use one precise word instead of rambling around. If anything this video shows me how not to do things.
@@shorgoth This becomes even worse when that grade level metric becomes a target. I had a class use that and it was a joke. Whenever my writing level is too low, just combine a few sentences with an "and" or other joining words to make it higher. That's not better writing. What a garbage metric.
Thanks. Is there a better metric to use?
There's a famous quote by Blaise Pascal - "If I had more time, I would have written a shorter letter." Generally, the ability to get your point across efficiently and effectively will result in a lower grade level.
As such, grade level as a metric is useful not to indicate writing quality, but rather literally just how complex the per-sentence structure is. This may very well be an indicator of how complicated the concepts are - scientific papers, for example, use big words and nuanced ideas. However, it's just as likely to be a result of poor planning on the part of the writer. Or maybe it's being done on purpose, for rhetorical or aesthetic purposes (ie, poetry.)
Speaking as a Programmer/Writer in Big Tech - I usually write around a 5th grade reading level. And that's for things like "how do I programmatically train an AI on Facebook-scale servers" or "how do I create a distributed E-to-E business application using the Windows ecosystem". This has nothing to do with the complexity of the subject, but rather a general rule of technical writing of "average sentence length no more than 20, max of 25."
Just to look at this particular example: Trump tends to ramble, while Biden has to work around his speech impediment. Neither is an indicator of intelligence or sophistication, but Trump's sentence structure is (likely) more complex due to his stream-of-consciousness speaking style, while Biden's is (likely) less complex due to his conscious choice to avoid stuttering.
(Yes, these are written speeches. The speech writers, if they're good, likely use the cadences their orators are most comfortable with.)
It is important to point out that speaking at a lower reading level is a sign of transparency and not an indication of the speaker's intellectual prowess. Speeches are written at a lower reading level in an effort to resonate with a wider group of viewers. In my work, I am asked to respond to customers at a fourth grade reading level.
Well said.
A guy at my work likes to use fancy rarely used words. Everyone has to look up his words just to figure out what he's asking. It's not an efficient form of communication and we usually get to his items last out of frustration.
True. Democrats have to speak at a lower reading level because their voter base has more blacks, hispanics and illegals, who by all objective measures have lower levels of education. That's why Bernie Sanders could never get his foot in the door. He was too verbose for his own voter base.
i think speaking at a lower reading level reflects the belief that the speaker has about the intelligence of who they are talking to. also i believe that a the majority of people cap out at about the grade reading level of movies and sitcoms plus the lingo required for any profession they work in.
@@toddspangler6669 lol ya that's not productive. some people say that birds of a feather flock together but on rare occasion some birds congregate in close proximity of those with different plumage.
The way you are measuring reading level might favour long rambling sentences that never go anywhere..
True. Which is why Trump's speeches were a higher "grade" level than Bidens.
The problem with all of this analysis, is that politicians, or at least very few, write their own speeches anymore. Political speeches these days, and especially presidential inauguration speeches, are written by professional speech writers, most of whom have a marketing background. They are no longer the thoughts and ideas of the elected, they are just advertising and brand management for the regime.
Presidents haven’t written their own speeches ever with few exceptions. Washington famously had Hamilton write his.
I don’t even think Trump reads a speech. He be rifting half the times and says what he feels. I like it.
The trump speech actually sounded fine and then I remembered reading this comment 😂
@@aqdadrIn his and many other presidents inauguration speeches, he used a teleprompter.
@@aqdadr He definitely does, but not for an innogural address. Which is smart. Those are historic, you dont wing those. But his rallies and conferences, yeah he definitely veers off at his leisure. Sometimes its great, sometimes it gets dicey....Mostly because we know if he farts, his haters will say he was playing with explosives around children.
You guys need more subscribers, this high quality and interesting content without at least 50k?
its one person
I / we thank you 😁
@@DataTime27 lmao!
@BasedArmenian-kh8pb Yes it is one person, look at the clone which isn't speaking and he'll be completely frozen a lot of the time, because it's just an editing trick.
FYI there is no "frozen" moment. I'm just sitting very still and motion is not perceivable compared to the noise on the camera
For the negativity, it could be influenced by war. In ww2, ppl said "we fight with guns and bombs", which could be counted as negative.
I was thinking economic
I think the up and downs of negativity in the pre-20th century likely can be explained by specific historical events that were happening. As noted later the most negative Republican speech was given in 1865 which, well... is late during the American Civil War and thus an increase in negativity would be expected after hundreds of thousands had been killed and the worst war America fought was slowly grinding to a close.
It impresses the heck out of me that there are people who can give facts without bias! Cheers Data Time.
AI (large language models) is not factual. 😂
@@mrtechie6810 It's more factual than your going to find on the TV news.
Statistical models like this have been around for a VERY long time and are widely accepted. 😂
@@datsuntoyy I don't even own a TV. So...
I feel like LLM sentiment analysis can be a *touch* iffy. Like the analysis itself can be biased by how the LLM was trained.
3:00 this also doesnt seem to quite get at the quality of the speech, though that is more quantifiable in an objective way. But there is something about being able to make your words succinct that is valuable. Which you kind of even demonstrated here.
12:26 interesting though, looking at the actual *contents* of the individual negative parts, Trump's speech gets at real deep seeded issues that have long term effects, things that we know are happening. Where Biden kind of just rags on a recent tragedy, and more emotionally charged things rather than things that are definitely happening.
That was my first thought. I wonder if an LLM can recognize its own bias.
This.
Trump's high grade level for his inaugural speech (as well as it is obvious to me when we have Teleprompter Trump vs. off the cuff Trump) is telling that it was reported that Trump's inaugural speech was written by Stephen Miller and Steve Bannon.
"was written by"
All presidential speeches are written by speech writers.
Steve bannon thought he was gonna be the next dick Chaney.
Trump's prepared speeches (notably his States of the Union) were significantly better than his usual speeches that usually had a more impromptu style. I wonder how the off the cuff speeches would score on the grade level scale. The words are fairly simple, but his sentences tend to be very long, even rambling.
@@ericwiddison7523 Good point. But with all the media lies, and Kamala Harris word salad, I'd prefer Trump any day. His results are far better than those of Biden, Obama, Bush, Clinton, and Bush.
😎♥✝🇺🇸💯
@@RodMartinJr I don't think you are paying attention. Trump's unscripted remarks are so disjointed that they would need to be cleaned up to qualify as word salad. He struggles to form complete sentences, much less coherent ideas.
I get that you don't like Harris's policies. That's the best basis for choosing how to cast your vote. But for you to claim that an unscripted Donald Trump is somehow more coherent a speaker than Harris or Biden is objectively false.
When someone I know was thinking about writing for the newspaper, their editor told them they couldn't go above an 8th grade reading level or no one would read what they wrote.
The "grade level" argument in the speeches probably reflects a similar intent on the part of the various presidents. They've definitely studied oratory, at least the last two; I recognized some of their techniques from my speech class.
You are correct, nevertheless it still holds true. The reading level is lower, either because it must be so or because it was contrived to be so.
@@waynehendricks1529 my point was that it is not that the speaker has a 5th grade command of English; rather that they've been advised to keep it simple.
On the other hand, the older they get, the smaller their vocabulary appears to become....
I guess it's a little of this, a little of that.... 🤷
The problem is the AI doesnt understand that Trump's speeches are written for him - he's just reading. Have it compare their interviews and press conferences instead.
It’s actually pretty much what Trump says in his rally. He repeats the same thing
@@eliza6722 No, not the rallies- the extemperaneous speaking when being questioned. Trump's and Harris'..... and Biden's.
The problem is the Biden very rarely holds press conferences or interviews. Remember when Trump would show up at the daily media briefing and get unscripted questions from the press? That's not a Biden thing.
@@peterbaruxis2511Trump often faces hostile interviewers, while Democrats only talk to allied media. Harris has never done an uncut unterview. So that will skew any data.
@@RenegadePlayingGames Of course, you're right but people eat this sh** up that they see on ABC, CNN etc.
One thing I would like to note though is "negativity" isn't always so easily measured because there is some negativity indicated in statements that are grammatically "positive". For example, "We will make America great again" indicates a loss of greatness prior to that.
I don't think a reference to the fact of an obvious loss of greatness qualifies as negativity. It's like... a matter of public record, man.
Have to laugh at myself for getting giddy at the prospect of watching this video! Being a bit of a data nerd, I found this video absolutely fascinating!!
This video is better than most as it describes using more than an LLM, but I guarantee you there are viewers who only hear "AI" and nothing else. These viewers don't understand nuance and may even think that ChatGPT can do perform such unbiased tasks.
Yes, AI is a singular monolith in unknowing minds. Much like 'genetically engineered' completely lacks nuance. Oh for decent education for all.
Remember when feeding criminal data to a AI to prove racial bias on the justice system and it concluded women are given lower sentences and there are too few blacks in jail lol
I see a lot of promise in this channel. The future is bright!
I think that you missed a key point which most miss. The right and left have different definitions or understandings of the words. They really are living in different realities.
The so-called left has decided to redefine en masse terms long understood by the masses. This is by design as is everything else they do and simply because they’re playing their supporters, using them as tools; a means to an end. The establishment right is guilty as well about to a lesser extent. The populist right by and large uses the established definitions. Sadly there’s no doubt all do exist in different realities,
Correct. Discernment is key..
My non-binary infant feels this is fascist.
This is fantastic! Could you do more analysis between speeches of Kamala Harris and Donald Trump? In particular I'm interested in seeing just their everyday speeches. And perhaps investigate the use of negative words, not just overall negative talk. Keep up the amazing work! You could write a dissertation on this material alone!
Thanks!
Yup. One is word salad and the other is what Trump is saying.
@shitfarmer8686 you actually have that backwards. Trump mentioned planes in the civil war in a speech once and rambles on about water pressure and being electrocuted by batteries in boats that are sinking. That can't happen unless you are actually touching the terminals and it's not powerful enough to kill you and would short circuit quickly in the water. He lies about statistics and caters to his crowd in speeches and lies based on who he is speaking to. Project 2025 was actually co-written by people he had in his administration and they have an agenda to create a theocratic dictatorship. Watch shiny happy people on Prime. The now married older Dugger girls from the TLC show 19 kids and counting did a 4 part docuseries on how they were abused and raised in their "quiverful" religion and the plans their religious leader has to infiltrate goverment positions. They will be the people Trump plans to replace thousands of civil servants with who will create chaos with forcing religion into job requirements and clog up the courts with challenges to it. He is a foreign agent of chaos to take down the U.S. from within as the authoritarian govts band together under the BRICS + currancy and want to change the power balance in the world for business and trade. I have been watching things closely for 8 years and I see how Trump increased the national debt by 7 trillion even before covid, and caused a downgrade in the U.S. credit rating after his government shutdown while in office. Took the country out of the human rights council and wanted to leave NATO...which would alienate the U.S. allies. Look at the big picture from a distance and it's very clear how he praises all those dictators and did some very strange foreign policy while in office that caused U.S.farmers to require support payments while giving permanent tax credits to the rich and the middle class which expire in 2025. That tax money could have helped pay for infrastructure or VA programs. It's all smoke and mirrors and they use social media algorithms to keep people in a rabbit hole of fake rage over drag queens or Disney turning them into voting pawns that vote against their own best interest.
One really simple analysis you can do is animated word clouds. Plot the word cloud for each speech as a whole, or after getting the llm to distill it down. Then animate between them over time.
this is so fuckin cool ugh and that intro was SO awesome. I also feel like this was super super accessible even if you knew nothing about ai
Thanks!
That's not the computer telling you XYZ it's the people who programmed it.
You don't program a LLM, you only choose what data to provide to it.
As a Psychiatrist I will tell you that this will never be conclusive because ai doesn't have a perception based on life experience or body language. There's also the intricacies I learned in Political psychology as well lol
As a software engineer, nobody wants to know it’s a similarity probability calculated based on the most likely next word.
This’ll be fun when the dataset is majority AI generated because then the bell curve center will just get higher as it submits it’s own circular logic.
No offense, a podiatrist should know that. No offense to either.
@@peterbaruxis2511 funny. Do you even know what a psychiatrist is? I am not a psychologist, huge difference. A doctorate in psychiatry and md license is much above a foot doctor. That isn't all of my studies either. So before you prove your ignorance and summon me to a battle of wits, I refuse because you're simply unarmed.
@@MatthewYoung-bq6rn So you don't think that a podiatrist, or an electrician or anyone without a doctorate in psychiatry and a medical degree understands that AI lacks perception based on life experience or body language? BTW, even Robert Young who wasn't really a doctor but just "played one on on tv" used the abbreviation MD. You've proven your ignorance- not only explicitly but implicitly. I don't have a PhD & I know you don't either. Ask AI or one of your peers if a psychiatrist is "much above" a foot doctor. Wits? Prove it.
Which party refused to show up for subpoenas?
Who rants about sharks and batteries?
Who rants about dogs and cats being eaten?
Love you guys! Watched the entire video
Something to perhaps consider when separating Democrats from Republicans is when they flipped ideologies & platforms over the period of time between the 1860’s and 1936.
LOl, oh, thats right...Thats what you guys say to try to distance yourselves from the fact that your party started the KKK and fought to keep slavery around. But I guess keeping people in section 8 housing dependent on barely enough to live on ISNT a form of slavery right? Or keeping the black community at only 13% of the population by pushing abortion is just a "coincidence"?
Now try dictators
It would be interesting to see if there are any native English speaking dictators who have published speeches that I can download
@@DataTime27 That sounds so oddly specific but it's actually what we would need in this situation
@@asmiledcrocodile LLMs are almost exclusively trained in English, hence the specificity.
@@nathan4678 Also things like grade-level and positivity are based on English sentence structures and words.
Excellent video boys! I think If you took Trump and Biden without a teleprompter, I'm pretty sure Trump's lead would only widen.
I figure the inaugural speech is the least important speech of their career. They already won. The stakes are too low. Nothing depends on it. In their first term, they have no experience to talk about. The main purpose seems to be blowing smoke up a select few asses, since practically no one listens to it. An analysis of campaign speeches would be better, but what I'd rather see is State of the Union speeches.
FDR sadly did not “serve four terms”, but he was elected four times.
Sadly he died in office.
FJB, absolute disgrace
Fdjt
Keep crying.
@@AA-pk6fo FDJTRUMP
@@Mrbb1996see you on the barricades, comrade
@roberthuff3122 I'm looking forward to, snow flake
Do Kamala
NO THANKS! Im 47 and STD free. I Intend to stay that way.
Not in a trillion years, and not for 17 trillion dollars.
How would you score a reading level based on cackling?
As a software developer, I will say that the AI is still subject to the bias of the person who made it. Im going to make a few guesses before watching, as a sort of experiment of my own.
Any mention of gun rights will probably be treated as negative, regardless of the context.
Historical famous speeches will likely be seen as more positive.
Left-leaning policies or ideas will probably be seen as "more positive" in general, because most AI models are made by left-leaning companies, and many of those ideas are reliant on appeals to emotion. Also, modern democrat speeches will probably be lower reading levels than older speeches in general.
That's the main guesses I'll make. Time to see how I did...
Okay, it looks like I was right about democrat speeches having lower reading levels now than most speeches in the past, but all the other guesses were not really covered in the video, so they didn't really matter. Oh well, still an interesting thought experiment.
I input 2 hours of unscripted speech by Trump and AI just said *_TILT!_*
AI doesn't know how to interpret sharks, electrocution and Hannibal Lector. 😂
The AI was sleeping on that last speech of Biden's so-called "positive" part of his speech. It was absolutely filled with negatives.
Yes, it listed a lot of negatives but the entire list was prefixed with "I ask every American to join me [...] uniting to fight the foes we face:" followed by that list of foes. In context the negative list was a target for positive outcomes encased in a "call to action" which is always considered a positive element in speechwriting.
@@LetYourLightShine5218 It was erroneous to count that as positive. Outlining the "foes we face" is clearly negative.
So, the reading level was determined by sentence length and syllables? tRump often rambles on and on with those long nonsensical sentences, would that not effect the score?
Trump was focused when he wrote and read his inauguration speech. Normally he's at a rally and speaking more freely. So he goes off on tangents or piles on adverbs and adjectives to get his talking points across, so it ends up being messy.
"Affect" the score.
Getting reading level from length of sentences and number of polysyllabic words could be problematic. If the AI failed to analyze for coherent thought, it would overestimate the reading level of gobbledygook and word salad.
Good vidoe
Trump's speeches are not his words, that's why he scored a higher grade level than Biden. Try the same with press conferences or rally speeches 🙂
Try the same in their most recent debate before Biden dropped out :)
It’s just when trump goes off script that you can hear how stable genius he is.
No modern president writes their own speeches. It would be better to analyze unscripted interviews. I think we'd get the same results though. Biden can barely string together three words without flubbing it.
@@angryanne he is way smarter than Komunala for sure.
bcs biden are his...
I will watch this entire video, later, however, I wanted to thank you for putting up two very good pictures in your thumbnail. It speaks volumes of your character.
Remember: positive or negative does not necessarily correlate to morally good or bad.
AI can be programmed to be just as biased as the people who wrote it. That’s too bad. Scientists should be completely neutral on subjective matters, like politics.