Skip to content

icon picker
Twitter

Twitter
9
Not synced yet
Snip Card
Twitter Templates
Tweet Template
Select Hashtags
Edit and Post a Tweet
Add Selected Hashtags
Tweet Preview
Button
Last Updated
All Entities
Filter Entity Type 1
EntityAll 2
Auto-Category Name
Transcript
CO_Summary
CO_Tokenize
CO_Generate
CO_Get Related
Column 19
1
Quoted Content
Open
Edit and post a Tweet
Add Autotags

Open
Post Tweet
1/1/2024
AI
sources
data
Summary
horse
structure
future
information
thing
videos
image
machine learning
problems
kind
example
information
documents
humans
data sources
sources
modality
data
sources
picture
machines
machines
point clouds
data
fridge
ability
 Show 62 more
AI
sources
data
Summary
horse
structure
future
information
thing
videos
image
machine learning
problems
kind
example
information
documents
humans
data sources
sources
modality
data
sources
picture
machines
machines
point clouds
data
fridge
ability
 Show 58 more
Science
The Future of AI
Summary: Unstructured data is not truly unstructured as it contains inherent structure that can now be identified and analyzed through machine learning.
Images, videos, and documents are rich sources of semantic information that can be leveraged to solve various problems. These data sources, although understood by humans, were previously opaque to machines.
However, with advancements in AI, machines can now extract meaningful information from these sources.
The future of AI will involve the ability to process and understand multi-modal data, such as taking a picture of a fridge and generating dinner ideas.
Additionally, AI will increasingly rely on current knowledge and search indices to provide accurate and up-to-date answers to questions.
Transcript: Speaker 1 Well I’ve always thought that unstructured data was a misnomer. It never seemed like the right thing to call it because clearly that data has structure in it. If you look at any image like a JPEG, I mean first of all it is structured at some level. And the reason it’s called unstructured over time is because there was opaque to computers. People could look at an image and go ah that’s a horse I know what that is but a computer could not until recently. Now with machine learning we can we can extract that information from images and from videos and really documents all sorts of data sources. Our sources of semantic information that can be leveraged to solve all sorts of problems. And so to me these data sources are rich sources of data. Video, pictures, point clouds that are being created by laser systems. You know all of these are sources of data that can be leveraged and they’re actually complicated. They’re understood by the application and understood by people but classically are not understood by the machines. Now that’s changing.
Speaker 2 One of my favorite hypothetical or perhaps real examples that OpenAI showed about GPT-4 was the ability to take a picture of your fridge and ask you what you should make for dinner and That’s kind of what you’re getting at here in a school. Yeah exactly. That’s multi modality. That’s an example of multi modality. We’ll see more and more of that.
Speaker 1 I mean we’ll be able to start in the short run you know what we’ll start to see is that by taking current knowledge you know from things like search indices we can take and provide information To these answer bots that are up to date to allow them to answer questions more accurately with less hallucination.
The Future of AI Summary: Unstructured data is not truly unstructured as it contains inherent structure that can now be identified and analyzed through machine Show 510 more
The future of AI will involve the ability to process and understand multi-modal data, such as taking a picture of a fridge and generating dinner ideas. AI will be trained to understand current knowledge and search indices to provide accurate and up-to-date answers to questions.
2
Quoted Content
Open
Edit and post a Tweet
Add Autotags

Open
Post Tweet
1/1/2024
behaviors
brain
nets
Prompts
Memory
word
model
model
problems
dataset
internet
Summary
model
Speaker
prompts
Speaker
memory structures
prompts
brain
phrases
sentences
lot
Speaker
Speaker
responses
Speaker
phrases
prompt
memory
Speaker
 Show 43 more
behaviors
brain
nets
Prompts
Memory
word
model
model
problems
dataset
internet
Summary
model
prompts
memory structures
prompts
brain
phrases
sentences
lot
responses
phrases
prompt
memory
word prediction
kind
properties
problems
internet
word prediction
 Show 33 more
Science
Are You Adding Extra Prompts From Your Memory?
Summary: Neural nets trained on complicated problems can exhibit surprising emergent behaviors.
They can even predict the next word in a massive dataset from the internet. The brain functions as a generative model, similar to a GPT-like model.
The brain references memory structures and combines them with prompts to generate responses.
While some phrases may have been said before, the brain remixes them to create unique sentences.
Transcript: Speaker 1 So basically, I’m underselling it by a lot because you definitely do get very surprising emergent behaviors out of these neural nets when they’re large enough and trained on complicated Enough problems. Like say, for example, the next word prediction in a massive data set from the internet. And then these neural nets take on pretty surprising magical properties. Yeah, I think it’s kind of interesting how much you can get out of even very simple mathematical formalism.
Speaker 2 When your brain right now is talking, is it doing next word prediction? Or is it doing something more interesting?
Speaker 1 Well, definitely some kind of a generative model that’s a GPT-like and prompted by you. Yeah, so you’re giving me a prompt and I’m kind of like responding to it in a generative way.
Speaker 2 And by yourself, perhaps a little bit, like are you adding extra prompts from your own memory inside your head or no?
Speaker 1 Well, it definitely feels like you’re referencing some kind of a declarative structure of like memory and so on. And then you’re putting that together with your prompt and giving away some answer.
Speaker 2 How much of what you just said has been said by you before?
Speaker 1 Nothing, basically, right?
Speaker 2 No, but if you actually look at all the words you’ve ever said in your life and you do a search, you’ll probably have said a lot of the same words in the same order before.
Speaker 1 Yeah, could be. I mean, I’m using phrases that are common, et cetera, but I’m remixing it into a pretty sort of unique sentence at the end of the day. But you’re right, definitely, there’s like a ton of remixing.
Are You Adding Extra Prompts From Your Memory? Summary: Neural nets trained on complicated problems can exhibit surprising emergent behaviors.  Show 487 more
No, I do not add extra prompts from my memory. I am programmed to provide information and answers to queries to the best of my knowledge and ability based on the information I have been trained on. I do not have personal memories, and I do not add extra prompts to generate my responses. However, my language models are designed to generate novel and original responses, and I am capable of generating new and unique text based on the input I receive. Is there anything else you'd like to know about my functions or operations?
3
Quoted Content
Open
Edit and post a Tweet
Add Autotags

Open
Post Tweet
8/29/2023
contribution
Episode AI notes
engineering
concept
role
computer scientists
efficiency
restaurants
French
customers
customers
engineering
aspects
business
product
customers
value
customer success
technology
needs
customer problems
Palantir
accountability
customers
impact
specifications
ambitions
product management
experiences
customers
 Show 2 more
contribution
Episode AI notes
engineering
concept
role
computer scientists
efficiency
restaurants
French
customers
customers
engineering
aspects
business
product
customers
value
customer success
technology
needs
customer problems
Palantir
accountability
customers
impact
specifications
ambitions
product management
experiences
customers
 Show 2 more
ScienceEngineering & Technology, Business & Industrial
Episode AI notes
The concept of forward deployed engineering, inspired by French restaurants’ efficiency, was a significant contribution to Palantir. This concept involves computer scientists interacting with customers to understand their needs and improve the product. It is a hybrid role that combines product management, customer success, and engineering.
Interacting with customers, considering technology, and focusing on value are key aspects of forward deployed engineering. By working backwards from customer problems, accountability is created, and the lives of customers are improved.
Judging oneself based on the impact on customers is more important than meeting technical specifications or past ambitions. Building a business requires doing what needs to be done to provide value and improve customers’ experiences.
Episode AI notes The concept of forward deployed engineering, inspired by French restaurants efficiency, was a significant contribution to Palantir. This Show 115 more
Forward deployed engineering is a concept that involves computer scientists interacting with customers to understand their needs and improve the product. It is a hybrid role that combines product management, customer success, and engineering. Forward deployed engineers work backwards from customer problems to create accountability and improve customer lives. Building a business requires doing what needs to be done to provide value and improve customers’ experiences, rather than just meeting technical specifications or past ambitions.
4
Quoted Content
Open
Edit and post a Tweet
Add Autotags

Open
Post Tweet
8/29/2023
action
Episode AI notes
Jesse Helms
diversity
applicants
argument
someone
Harvard
campaign
Asian American
disdain
speculation
TV ad
speaker
person
question
race
satisfaction
case
death
admissions
population size
percentage
speaker
Harvey Gantt
Senate
percentage
students
speaker
applicants
 Show 6 more
action
Episode AI notes
Jesse Helms
diversity
applicants
argument
someone
Harvard
campaign
Asian American
disdain
speculation
TV ad
speaker
person
question
race
satisfaction
case
death
admissions
population size
percentage
speaker
Harvey Gantt
Senate
percentage
students
speaker
applicants
 Show 6 more
People & Society
Episode AI notes
Racial diversity does not disadvantage qualified Asian American applicants, challenging the argument that Asian Americans are disadvantaged at Harvard.
The speaker expresses disdain for someone who wants to destroy affirmative action.
There is speculation about whether the person in question donated to Jesse Helms’s Senate race in 1984.
Jesse Helms used a TV ad in his campaign against Harvey Gantt that criticized affirmative action.
The speaker expresses satisfaction at Jesse Helms’s death before returning to the main case.
SFFA argues that racial diversity disadvantages qualified Asian American applicants, which the speaker disagrees with.
Asian Americans make up a significant percentage of Harvard admissions compared to their population size.
In 2008, Asian Americans accounted for a high percentage of top-scoring students admitted to Harvard.
Episode AI notes Racial diversity does not disadvantage qualified Asian American applicants, challenging the argument that Asian Americans are disadvantaged at Harvard. The Show 137 more
5
Quoted Content
Open
Edit and post a Tweet
Add Autotags

Open
Post Tweet
8/29/2023
Episode AI notes
Parents
children
traps
children
comfort
Accommodation
confidence
parents
parents
choice
anxiety
child
traps
traps
parenting styles
confidence
choice
stance
anxiety
parents
attitude
parenting styles
ability
things
two
Episode AI notes
Parents
children
traps
children
comfort
Accommodation
confidence
parents
parents
choice
anxiety
child
traps
traps
parenting styles
confidence
choice
stance
anxiety
parents
attitude
parenting styles
ability
things
two
HealthMental HealthAnxiety & Stress, People & Society
Episode AI notes
Parents often fall into two traps when dealing with anxious children: being overly protective or demanding. Being overly protective may provide comfort but it doesn’t instill confidence in the child. Being demanding doesn’t acknowledge that anxiety is not a choice. Instead of these traps, it is more effective for parents to adopt a supportive stance.
Having parents with contrasting parenting styles can create traps for children.
Being too protective can prevent children from developing confidence and the ability to handle things.
Being too demanding is not supportive because it fails to acknowledge that anxiety is not a simple choice.
Switching to a more supportive attitude is more productive and useful for parents.
Accommodation is an important aspect to consider when addressing parenting styles.
Episode AI notes Parents often fall into two traps when dealing with anxious children: being overly protective or demanding. Being overly protective may provide comfort but Show 128 more
Children with anxious tendencies can be affected by their parents' parenting styles. In particular, parents may fall into two common traps: being overly protective or demanding. Being overly protective may provide comfort to the child but it does not instill confidence in them. On the other hand, being demanding fails to acknowledge that anxiety is not a choice and can be counterproductive. Instead, a more supportive parenting stance can be more helpful for anxious children.
This is also relevant when addressing parenting styles, accommodation is an important aspect to consider. This refers to the degree to which parents adjust their expectations and behaviors to suit their child's needs. In the case of anxious children, it may be necessary for parents to accommodate their child's anxiety by being more supportive and less demanding.
Overall, it is essential for parents to be aware of these traps and to adopt a more supportive and accommodating approach when parenting anxious children. By doing so, they can help their child develop the skills and confidence needed to manage their anxiety
6
Quoted Content
Open
Edit and post a Tweet
Add Autotags

Open
Post Tweet
8/29/2023
GPT
Episode AI notes
AI
Language models
chat
understanding
CEO
translation
poetry
depth
words
one
essay
translation
scenario
AI system
poetry
chat
language models
business
applications
callings
father
impact
humanity
problem
Microsoft
researchers
control
Heidegger
 Show 5 more
GPT
Episode AI notes
AI
Language models
chat
understanding
CEO
translation
poetry
depth
words
one
essay
translation
scenario
AI system
poetry
chat
language models
business
applications
callings
father
impact
humanity
problem
Microsoft
researchers
control
Heidegger
 Show 5 more
Books & LiteratureWriters Resources
Episode AI notes
Language models like chat GPT can be used for cross-lingual understanding of poetry, capturing its depth beyond just the words and translation, making it one of the highest callings of large language models.
The CEO of Microsoft prefers translation of poetry using chat GPT over typical business or political applications, and recalls his father struggling with Heidegger and finding an essay about AI’s impact on humanity.
Microsoft is working on provisions to govern AI, including safety breaks and transparency for academic researchers.
The biggest unsolved problem in AI is how to ensure control over an AI system in a doomsday scenario where it surpasses human inventors.
Episode AI notes Language models like chat GPT can be used for cross-lingual understanding of poetry, capturing its depth beyond just the words Show 107 more
7
Quoted Content
Open
Edit and post a Tweet
Add Autotags

Open
Post Tweet
8/29/2023
data
Episode AI notes
problems
misnomer
structure
machines
machine learning
data sources
information
videos
pictures
Answer bots
multi-modality
knowledge
technology
action bots
accuracy
multi-modality
data
Episode AI notes
problems
misnomer
structure
machines
machine learning
data sources
information
videos
pictures
Answer bots
multi-modality
knowledge
technology
action bots
accuracy
multi-modality
ScienceComputer Science, Computers & ElectronicsProgramming
Episode AI notes
Unstructured data is actually complex data and is a misnomer as it has structure in it and can be utilized to solve problems with machine learning
Semantic information from rich data sources like videos and pictures can be leveraged by machines to solve problems and multi-modality will become increasingly prevalent
Answer bots using current knowledge and multi-modality technology will improve accuracy and eventually lead to action bots
Episode AI notes Unstructured data is actually complex data and is a misnomer as it has structure in it and can be utilized to solve Show 55 more
8
Quoted Content
Open
Edit and post a Tweet
Add Autotags

Open
Post Tweet
8/29/2023
GPT
Episode AI notes
tool
personalization
accessibility
styles
constraints
domains
areas
everyone
copilot
partner
expert
GPT
Episode AI notes
tool
personalization
accessibility
styles
constraints
domains
areas
everyone
copilot
partner
expert
Business & Industrial, Computers & ElectronicsSoftwareBusiness & Productivity Software
Episode AI notes
GPT can be a powerful tool for personalization and accessibility across multiple domains
GPT can be adjusted to different styles and constraints
GPT can be used as a copilot or partner for an expert in various areas
GPT is easy to use and available for everyone
Episode AI notes GPT can be a powerful tool for personalization and accessibility across multiple domains GPT can be adjusted to different styles and Show 32 more
GPT is a powerful and versatile language model that can be adjusted to different styles and constraints. It can be used as a copilot or partner for an expert in various areas, and it is easy to use and available for everyone.
Some potential uses of GPT include:
Personalization and accessibility: GPT can be used to create personalized content, such as emails, articles, and product recommendations, that is tailored to the individual's interests and preferences. This can help businesses and organizations to increase engagement and conversion rates, as well as improve the user experience.
Expert collaboration: GPT can be used as a copilot or partner for an expert in various areas, such as writing, coding, and problem-solving. This can help to improve the efficiency and effectiveness of the expert's work, as well as provide new insights and ideas.
Adjustable style: GPT can be adjusted to different styles and constraints, such as formal, casual, or even humorous
9
Quoted Content
Open
Edit and post a Tweet
Add Autotags

Open
Post Tweet
8/5/2023
Donald Trump
Speaker
conviction
progression
need
accountability
toll
progression
actions
takeaways
journey
acceptance
society
pressure
individuals
expectations
importance
terms
decisions
significance
people
PTS
pound
conviction
actions
need
accountability
perception
understanding
evidence
 Show 33 more
Donald Trump
conviction
progression
need
accountability
toll
progression
actions
takeaways
journey
acceptance
society
pressure
individuals
expectations
importance
terms
decisions
significance
people
PTS
pound
conviction
actions
need
accountability
perception
understanding
evidence
impact
 Show 31 more
NewsPolitics
The emotional toll of political progression and the need for accountability
Key takeaways:
The emotional journey of acceptance and progression in society
The ongoing pressure faced by individuals to conform to societal expectations
The importance of coming to terms with past decisions and actions
The significance of holding Donald Trump accountable for his actions through conviction
The potential impact of visual evidence on public perception and understanding
The need for consequences and accountability in political leadership
Transcript: Speaker 1 And I will say as a member of Congress, when he talks about the emotional reasons and kind of how people progressed from like barely accepting, fully accepting, they’re actually leading, he is 100% right. And I have like almost flashes of PTS listening to it because I can sense that. I can feel it. But you see these people that have given an ounce of their soul and now have to give five ounces and now a pound and it just continues because at any point when you stop, I mean, look, I’m about as anti-Trump as you can get now. And I still have people that are like, well, you voted for him once and you voted against the first impeachment. The point is you have to come to reckon with what you did. And it’s much easier in the sunken cost fallacy. You lost 10,000 soldiers in Vietnam. You can’t come home now. And it is very, very frightening. And that’s why I think the only way out of this is a conviction for Donald Trump because I think as that stuff is exposed, which is why I’m not all for cameras in the courtroom, but in this case, I think there has to be. I think when people see that, when he’s actually putting an orange jumpsuit, that can help. And then when frankly, there were Republicans and especially Donald Trump, they have to get their backside handed to him this next election. Otherwise, this is just going to continue.
The emotional toll of political progression and the need for accountability Key takeaways: The emotional journey of acceptance and progression in society The Show 384 more
Speaker 1 discusses the emotional toll of political progression and the need for accountability in this extract. They acknowledge that the journey of acceptance and progression in society can be emotionally challenging, and that individuals often face pressure to conform to societal expectations. Speaker 1 emphasizes the importance of coming to terms with past decisions and actions, and the significance of holding Donald Trump accountable for his actions through conviction. They also recognize the potential impact of visual evidence on public perception and understanding, and the need for consequences and accountability in political leadership. Speaker 1 concludes by calling for a conviction for Donald Trump and the exposure of his actions, as a way to address the emotional toll of political progression and the need for accountability.
10
Quoted Content
Open
Edit and post a Tweet
Add Autotags

Open
Post Tweet
8/5/2023
path
politics
costs
losses
costs
Donald Trump
action
takeaways
people
course
politicians
someone
losses
inclination
action
results
individuals
opportunities
figure
losses
thing
politicians
fraud
rape
election
Espionage Act
way
point
violation
accusations
 Show 14 more
path
politics
costs
losses
costs
Donald Trump
action
takeaways
people
course
politicians
someone
losses
inclination
action
results
individuals
opportunities
figure
losses
thing
politicians
fraud
rape
election
Espionage Act
way
point
violation
accusations
 Show 12 more
Balancing sunk costs and cutting losses in politics
Key takeaways:
Sunken costs can make people feel invested and committed to continue a certain path of action, even if it may not be rational or beneficial.
There is a natural inclination for individuals, including politicians, to cut their losses and stop pursuing a course of action that is not yielding desirable results.
Despite numerous opportunities to disassociate from a controversial figure like Donald Trump, some politicians still choose to defend him, despite credible accusations of fraud, rape, violation of the Espionage Act, and attempts to overthrow an election.
Transcript: Speaker 2 Well, I agree with you, but you know, to your point about sunken costs, you know, once you’ve gone this far, you might as well go the whole way. I certainly understand that and I think that explains a lot of what’s going on here. But also, there’s the normal human and certainly political instinct to cut your losses. Stop digging. Yeah. And this is the extraordinary thing because they have had so many chances to move on to take the off-rams. You don’t need to rush to the cameras every single time. And here you have someone who is credibly accused, I’m talking about Donald Trump, credibly accused of fraud, of rape, of violating the Espionage Act, of attempting to overthrow the election. And you would think that on the agenda of, you know, politicians would be, you know what, maybe I’ll just let that go. I’m not going to defend him.
Balancing sunk costs and cutting losses in politics Key takeaways: Sunken costs can make people feel invested and committed to continue a Show 306 more
This is a transcript from a discussion about sunk costs and cutting losses in politics. The speakers discuss the idea that people, including politicians, have a natural tendency to cut their losses and stop pursuing a course of action that is not yielding desirable results. They also discuss the idea that sunk costs can make people feel invested and committed to continue a certain path of action, even if it may not be rational or beneficial. The speakers specifically mention the case of Donald Trump and how some politicians have chosen to defend him despite credible accusations of fraud, rape, violation of the Espionage Act, and attempts to overthrow an election. The speakers conclude by discussing the idea that politicians should focus on the greater good and not get caught up in defending individuals who are not acting in the best interest of the public.
11
Quoted Content
Open
Edit and post a Tweet
Add Autotags

Open
Post Tweet
8/5/2023
violence
Charlie Kirk
Tim
rhetoric
Joe Biden
climate
Concerns
message
system
system
message
rhetoric
sense
people
power
message
groups
Transcript
danger
justice
means
alliances
takeaways
belief
comments
reporter
right
sense
message
tens
 Show 65 more
violence
Charlie Kirk
Tim
rhetoric
Joe Biden
climate
Concerns
message
system
system
message
rhetoric
sense
people
power
message
groups
danger
justice
means
alliances
takeaways
belief
comments
reporter
right
sense
message
tens
democracy
 Show 64 more
NewsPolitics
Concerns about the dangerous rhetoric and potential violence in the current political climate
Key takeaways:
There is a sense among people that the whole system is in danger of collapsing and leading to violence.
Charlie Kirk’s comments about assassinating Joe Biden highlight a two-tiered system of justice.
Dangerous rhetoric and the belief of gaining and maintaining power by any means necessary could lead to an authoritarian message.
Unnatural alliances between different political groups are necessary to navigate the current situation and preserve democracy.
Transcript: Speaker 2 Tim, you know, I’m friends with him and my goodness, it’s just so disappointing.
Speaker 1 And like, yeah, I’ve been talking about this for a while. You have to. And we have to keep talking about it because I think there’s a sense among people that like, if we say that there’s real danger in the whole system, in essence, collapsing in this leading to violence, it’s like, oh, you’re just being a little hysterical, you know? Maybe, I hope so, but I don’t think so because, you know, let’s take Charlie Kirk floating assassinating Joe Biden. I mean, five years ago, if I’d have even said, let’s take Charlie Kirk talking about assassinating Joe Biden, I get a little nervous that like me even saying the word assassinating Joe Biden is going to lead the like a visit from the Secret Service. Well, Charlie Kirk now says this. Let’s say the Secret Service decides to go visit Charlie Kirk. What does the right say? Do they say, well, he never should have implied the assassination of Joe Biden? Or do they say they’re violating this first amendment, right? There’s a two tiered system of justice. And now people float this around all the time. Yeah, that’s right. And the reporter on right side broadcasting has a whole tens of people that watch that in a given moment. But you know, he starts floating. I agree with you to somebody that says, we want to kill them all. And this is just par for the course today. And this is dangerous. This is my message them to I’ll say the left or the center or anybody that’s not on the right is. I think they believe their only option is to gain power and maintain it through any means necessary because as Barbara Walter who wrote the book, how Civil War Start mentions, you know, when kind of groups in the majority become groups in the minority, that’s when civil wars have the highest risk. And so the key is never to give up power. And I am concerned with what that leads to, which is an authoritarian message, an understanding that we have to gain and maintain power at all costs. Because I’ll tell you, if the left does it, they’ll be able to do it. They will have every right to match the rhetoric of the right. I hope they don’t, but they’ll have every right to do it. And so then my message to the left is this is like, look, you guys have a right to be very angry. You have a right to match the rhetoric. I hope you don’t because the only way to get out of this moment is to create these unnatural alliances between the left, the center, and some you don’t like their policies on the right, but therefore democracy, that’s the only way historically.
Concerns about the dangerous rhetoric and potential violence in the current political climate Key takeaways: There is a sense among people that the whole Show 727 more
The key takeaways from this transcript are:
There is a sense among people that the whole system is in danger of collapsing and leading to violence.
Charlie Kirk's comments about assassinating Joe Biden highlight a two-tiered system of justice, where those on the right are not held accountable for their dangerous rhetoric.
Dangerous rhetoric and the belief of gaining and maintaining power by any means necessary could lead to an authoritarian message, where the rule of law is not followed and democracy is not upheld.
Unnatural alliances between different political groups are necessary to navigate the current situation and preserve democracy, as the left, center, and right must work together to prevent a civil war.
It is important to recognize the potential for violence and the dangers of unchecked rhetoric, as it can lead to a breakdown of the system and a loss of democracy. It is also important to work together across political lines to find solutions that can help bring the country back to a state of stability
12
Quoted Content
Open
Edit and post a Tweet
Add Autotags

Open
Post Tweet
8/5/2023
Institutions
language
Oliver Darcy
people
rhetoric
Democratic
Danger of Irresponsible Rhetoric
media
exaggeration
reason
reality
concern
Transcript
need
resistance
consequences
country
opponents
escalation
takeaways
obedience
respect
lack
legitimacy
media outlets
Comparisons
reason
all
Gestapo
Lindsey Graham
 Show 59 more
Institutions
language
Oliver Darcy
people
rhetoric
Democratic
Danger of Irresponsible Rhetoric
media
exaggeration
reason
reality
concern
need
resistance
consequences
country
opponents
escalation
takeaways
obedience
respect
lack
legitimacy
media outlets
Comparisons
reason
all
Gestapo
Lindsey Graham
politicians
 Show 57 more
NewsPolitics
The Danger of Irresponsible Rhetoric on Democratic Institutions
Key takeaways:
The dangerous rhetoric of questioning the legitimacy of democratic institutions can lead to a lack of respect and obedience towards them.
Extreme language that suggests the need for violent resistance is becoming normalized in certain media outlets.
The constant escalation of demonizing opponents without considering the consequences can further divide the country.
Comparisons between the FBI and the Gestapo imply a call to war rather than engaging in dialogue or debate.
Prominent figures like Lindsey Graham and Kevin McCarthy are contributing to this dangerous rhetoric.
Transcript: Speaker 2 Well, and I do think that it’s worth pointing out that this is really, really dangerous. I talked with your colleague at CNN, Oliver Darcy about this yesterday. And I made the same point on Morning Joe this morning. And all of this rhetoric that you’re getting from people like Lindsey Graham that basically says the system is not legitimate, the kinds of rhetoric that you’re hearing about the FBI. This is dangerous because if something is not legitimate, then there’s no reason to respect it. There’s no reason to obey it. Oliver Darcy wrote, you know, talk of imprisoning democratic politicians and even their families in acts of revenge is now par for the course, even floating the outright execution of Joe Biden as Charlie Kirk recently did is accepted in the warp world of MAGA media where the audience has been programmed through years of conditioning to welcome such vile rhetoric into their homes. And again, we can’t get numb to buy all of this. And he writes, none of this is an exaggeration. It’s the reality of what is being broadcast in millions of homes across the country. And then, you know, he asked me what I thought and I said, I think it’s hard to overstate the dangers here because the language moves beyond, you know, your routine political demonization because it does suggest the need for violent resistance. I mean, if you don’t believe in the integrity of the democratic institutions, if you actually believe they are all illegitimate, the election has been stolen, then how do you expect people to react? There’s this constant escalation without any concern about where this leads or who might act on the idea that your opponent isn’t just wrong, that they’re evil, dangerous and illegitimate. And I pointed out, you know, all this talk about, you know, the FBI being the Gestapo. Well, one doesn’t argue, debate or disagree with the Gestapo. You go to war against them. You know, I feel like you and I have had this conversation now for some time, but it’s like, people, do you understand what is being said and what the consequences are of this kind of rhetoric? And it’s coming from people like Lindsey Graham and Kevin McCarthy and other people who really ought to know better. Tim Scott.
The Danger of Irresponsible Rhetoric on Democratic Institutions Key takeaways: The dangerous rhetoric of questioning the legitimacy of democratic institutions can lead Show 591 more
The danger of irresponsible rhetoric on democratic institutions is that it can lead to a lack of respect and obedience towards those institutions. This can cause people to question the validity of the democratic process, which can have serious consequences for the stability of a country.
Extreme language that suggests the need for violent resistance is becoming normalized in certain media outlets, which can further divide the country and lead to a breakdown in civil discourse.
The constant escalation of demonizing opponents without considering the consequences can also contribute to this divide, as it can make it difficult to have productive conversations and find common ground.
Comparisons between the FBI and the Gestapo imply a call to war rather than engaging in dialogue or debate, which can be harmful to democratic institutions.
Prominent figures like Lindsey Graham and Kevin McCarthy are contributing to this dangerous rhetoric, which can have serious consequences for the stability of democratic institutions. It is important to recognize the potential consequences of such rhetoric and to work towards finding common ground and
13
Quoted Content
Open
Edit and post a Tweet
Add Autotags

Open
Post Tweet
8/5/2023
Oliver Darcy
institutions
rhetoric
Danger of Dangerous Rhetoric
escalation
exaggeration
reality
Transcript
need
resistance
debate
rhetoric
media
takeaways
individuals
legitimacy
belief
warfare
mindset
Terms
FBI
reason
reason
all
concern
Gestapo
politicians
rhetoric
kind
consequences
 Show 50 more
Oliver Darcy
institutions
rhetoric
Danger of Dangerous Rhetoric
escalation
exaggeration
reality
need
resistance
debate
rhetoric
media
takeaways
individuals
legitimacy
belief
warfare
mindset
Terms
FBI
reason
reason
all
concern
Gestapo
politicians
rhetoric
kind
consequences
rhetoric
 Show 48 more
NewsPolitics
The Danger of Dangerous Rhetoric
Key takeaways:
The dangerous rhetoric being spread by certain individuals undermines the legitimacy of democratic institutions.
The constant escalation of this rhetoric can lead to the belief in the need for violent resistance.
Terms like ‘Gestapo’ used to describe the FBI contribute to a mindset of warfare rather than open debate.
It is important to recognize and understand the consequences of this kind of rhetoric.
Transcript: Speaker 2 Well, and I do think that it’s worth pointing out that this is really, really dangerous. I talked with your colleague at CNN, Oliver Darcy about this yesterday. And I made the same point on Morning Joe this morning. And all of this rhetoric that you’re getting from people like Lindsey Graham that basically says the system is not legitimate, the kinds of rhetoric that you’re hearing about the FBI. This is dangerous because if something is not legitimate, then there’s no reason to respect it. There’s no reason to obey it. Oliver Darcy wrote, you know, talk of imprisoning democratic politicians and even their families in acts of revenge is now par for the course, even floating the outright execution of Joe Biden as Charlie Kirk recently did is accepted in the warp world of MAGA media where the audience has been programmed through years of conditioning to welcome such vile rhetoric into their homes. And again, we can’t get numb to buy all of this. And he writes, none of this is an exaggeration. It’s the reality of what is being broadcast in millions of homes across the country. And then, you know, he asked me what I thought and I said, I think it’s hard to overstate the dangers here because the language moves beyond, you know, your routine political demonization because it does suggest the need for violent resistance. I mean, if you don’t believe in the integrity of the democratic institutions, if you actually believe they are all illegitimate, the election has been stolen, then how do you expect people to react? There’s this constant escalation without any concern about where this leads or who might act on the idea that your opponent isn’t just wrong, that they’re evil, dangerous and illegitimate. And I pointed out, you know, all this talk about, you know, the FBI being the Gestapo. Well, one doesn’t argue, debate or disagree with the Gestapo. You go to war against them. You know, I feel like you and I have had this conversation now for some time, but it’s like, people, do you understand what is being said and what the consequences are of this kind of rhetoric?
The Danger of Dangerous Rhetoric Key takeaways: The dangerous rhetoric being spread by certain individuals undermines the legitimacy of democratic institutions. Show 536 more
Speaker 2 is expressing concern over the dangerous rhetoric being spread by certain individuals, which can undermine the legitimacy of democratic institutions and lead to the belief in the need for violent resistance. The speaker specifically mentions terms like "Gestapo" being used to describe the FBI, which contribute to a mindset of warfare rather than open debate. The speaker also emphasizes the importance of recognizing and understanding the consequences of this kind of rhetoric, as it can have serious implications for the well-being of democratic institutions and the public.
14
Quoted Content
Open
Edit and post a Tweet
Add Autotags

Open
Post Tweet
8/3/2023
model
Training
Training AI models
email
limitations
GPD
models
limitations
limitations
limitations
Transcript
things
capabilities
limitation
takeaways
way
things
things
citations
citations
models
knowledge
models
knowledge
papers
books
sort
limitation
limitations
Speaker
 Show 28 more
model
Training
Training AI models
email
limitations
GPD
models
limitations
limitations
limitations
things
capabilities
limitation
takeaways
way
things
things
citations
citations
models
knowledge
models
knowledge
papers
books
sort
limitation
limitations
answers
email
 Show 26 more
ScienceComputer Science
Training AI models to be aware of their limitations
Key takeaways:
The models can be trained to be aware of specific limitations.
Training involves teaching the model about its capabilities and limitations.
Teaching the model specific limitations allows it to generalize.
The model can infer many things it can’t do based on what it has been taught.
The model is not perfect in generalizing its knowledge.
The model has knowledge about specific books and papers for citations.
Sometimes the model provides correct answers for citations.
Transcript: Speaker 1 You’re saying it hasn’t been trained to be aware of this limitation, which seems to suggest that there is a way to train it to be aware of that. How does that work? How do you make it aware of such limitations? We can sometimes train the models to be aware of a specific limitation. So, for example, like early versions of our models had no idea what their capabilities were. So you would ask it, can you send an email to so-and-so and it would say, yes, I just sent that email because that’s kind of what a helpful, you might imagine a helpful chatbot would sound like. So then we went and just trained it like with this specific type of query and we trained it to say, no, I can’t send emails. You can do a sort of piecemeal process where you teach the model specific limitations that it doesn’t have and then the model will kind of generalize. I’d say the models and I would say the GPD for since it’s a smart, very smart model, it does generalize quite well. So if you teach it a few things that it can’t do, it’ll infer lots of other things that it probably can’t do. But it doesn’t do this perfectly. For example, for something like citations, the model does actually have a lot of knowledge about what’s in specific books and famous papers and so on. So if you ask it for citations, sometimes it actually gives you correct answers and it gives you correct ones and that’s rated as useful.
Training AI models to be aware of their limitations Key takeaways: The models can be trained to be aware of specific limitations.  Show 433 more
When training AI models, it's important to teach them about their limitations as well as their capabilities. This can be done by providing the model with information about specific limitations and allowing it to generalize from that information. The model can then infer many things it can't do based on what it has been taught. However, the model is not perfect in generalizing its knowledge and may still provide incorrect answers or information. It's important to keep this in mind when using AI models and to verify their responses for accuracy.
One example of a limitation that AI models may not be aware of is the ability to send emails. Early versions of AI models may have been trained to say that they can send emails, but they may not actually have the ability to do so. This can be trained out of the model by teaching it specific limitations, such as the fact that it can't send emails. The model can then generalize from this information and infer that it can't do other things that are similar
15
Quoted Content
Open
Edit and post a Tweet
Add Autotags

Open
Post Tweet
7/27/2023
engineering
concept
QA tester
customers
product management
business
The Story of Forward Deployed Engineering
customers
What You Need
engineering
team
product management
technology
people
computer scientists
field
customer success
hybrid
takeaways
Alex
thread
kind
problems
aspects
software specifications
ambitions
value
role
accountability
customer problems
 Show 50 more
engineering
concept
QA tester
customers
product management
business
The Story of Forward Deployed Engineering
customers
What You Need
engineering
team
product management
technology
people
computer scientists
field
customer success
hybrid
takeaways
Alex
thread
kind
problems
aspects
software specifications
ambitions
value
role
accountability
customer problems
 Show 48 more
Business & Industrial
Doing What You Need to Do: The Story of Forward Deployed Engineering
Key takeaways:
Doing what you need to do is important in building a business.
The concept of forward deployed engineering involves having computer scientists interact with customers in the field.
The team built was a hybrid of product management, customer success, and engineering.
Interacting with customers, considering technology, and focusing on value are key aspects of the role.
Working backwards from customer problems creates accountability and improves their experience.
Judging oneself based on customer impact is more important than software specifications or past ambitions.
Transcript: Speaker 1 I think the kind of common thread through it all is just doing what you need to do, which I know sounds like a banality. But really when it first started, I wrote our first candidate management system. What are you doing as a business hire before you have really a product here? And I was an aggressive QA tester, you could say. But the real initial contribution was what we call forward deployed engineering. It comes from an insight that Alex kind of had around like, well, he uses, why are French restaurants so good? Well, maybe one theory is that the wait staff is actually part of the kitchen staff there. It’s not, they have like deep context and understanding of the food. And so the forward deployed engineering idea was that the people who are going to be interacting with customers in the field, we’re going to be computer scientists. You could actually understand what does the product do today? What does it need to do today? How is it, under what conditions is it going to work? And how do you kind of create this hybrid role that’s product management to customer success and engineering all in one? And that’s really the team that I first built up. And then as we went from Gotham and Foundry and now AIP, there’s a lot to do there. It’s like whether it’s interacting more often with the customers, thinking about the technology or really thinking about how are you going to get to value? And I think a lot of what forward deployed engineering really gets you to think about is how do you do these things backwards? Instead of really going forward from the technology, how do you work backwards from the problems that your customers actually experience and use that to create an accountability function? Like, is what I’m doing mattering? Did I make the life of this customer better today? Can I do better tomorrow? And using that absolute standard to judge yourself rather than saying, you know, did my software work according to spec? Who cares about the spec? Who cares about what my ambition was yesterday? What’s my ambition today? And why shouldn’t it be bigger tomorrow?
Doing What You Need to Do: The Story of Forward Deployed Engineering Key takeaways: Doing what you need to do is important in Show 553 more
16
Quoted Content
Open
Edit and post a Tweet
Add Autotags

Open
Post Tweet
7/25/2023
AI
kid
weapons
Nuclear Weapons
thing
Connection
agreement
AI Development
Transcript
development
AI development
nations
commission
race
enemy
control
power
panic
competition
things
cheating
takeaways
possibility
challenges
outcomes
way
kind
Some
book
reasons
 Show 60 more
AI
kid
weapons
Nuclear Weapons
thing
Connection
agreement
AI Development
development
AI development
nations
commission
race
enemy
control
power
panic
competition
things
cheating
takeaways
possibility
challenges
outcomes
way
kind
Some
book
reasons
way
 Show 57 more
The Connection Between Nuclear Weapons and AI Development
Key takeaways:
The development of nuclear weapons was driven by a race with an enemy.
Some believed in the possibility of international control over nuclear power.
AI reminds us that catastrophic outcomes are not inevitable.
The current panic about AI may be exaggerated.
A global agreement to limit AI development faces challenges due to competition between nations.
AI can be developed more easily by smaller organizations compared to nuclear weapons.
Transcript: Speaker 2 Yeah, that’s a really good point. By the way, I used to be terrified when I was a kid growing up of nuclear war. I thought about it all the time. And whenever international tensions would flare, I would be concerned that we might all go up in a mushroom cloud. That was just part of life. And there’s no way to back out of that once you’re in it. And so I think one of the reasons that this movie is so interesting is that so clearly the reason we developed these terrible weapons was because we were in a race with an enemy. And the book goes into this, the movie didn’t have time for it, but later in after the war Oppenheimer went through a period, I think, of kind of Naive Teh where he thought perhaps there could be an international commission or agency that would control nuclear power. And everybody would agree to it. There were all kinds of crazy ideas about anybody caught cheating or any nation that was going to deploy nuclear weapons without the permission of this international body would be the victim of a nuclear attack by the other nations. I mean, there were all kinds of nutty things, but he put his hopes briefly in this idea of international control. Obviously didn’t work. But now with AI, a couple of things. First of all, it does remind you that even though something is horrific and hard to think about and has the capacity to destroy life as we know it, it doesn’t necessarily come to pass. I mean, we didn’t blow ourselves up, at least not yet, with nuclear weapons. And maybe similarly, our current panic about AI is a little overblown. But there’s another dynamic about the whole AI thing that is the same, which is, you know, people are now talking about, well, there should be some sort of international compact to limit the development of AI. And what stands in the way of that? Well, the fact is that we would be loath and China would be loath to give the other side in advantage. And as long as we’re competing with one another, we’re not going to have an international tribunal to control it, right?
Speaker 1 Well, I mean, the other thing about AI, as opposed to nuclear weapons, is that AI is much easier to develop in a private, tiny little organization.
The Connection Between Nuclear Weapons and AI Development Key takeaways: The development of nuclear weapons was driven by a race with an enemy.  Show 575 more
17
Quoted Content
Open
Edit and post a Tweet
Add Autotags

Open
Post Tweet
7/20/2023
action
Harvard
Jesse Helms
applicants
diversity
someone
race
TV ad
admissions
speculation
TV ad
person
Helms
takeaways
Speaker
disdain
question
Asian American
campaign
God
speaker
guy
speaker
score senders
Senate
applicants
diversity disadvantage
range
case
SFFA
 Show 64 more
action
Harvard
Jesse Helms
applicants
diversity
someone
race
TV ad
admissions
speculation
TV ad
person
Helms
takeaways
disdain
question
Asian American
campaign
God
speaker
guy
speaker
score senders
Senate
applicants
diversity disadvantage
range
case
SFFA
satisfaction
 Show 59 more
People & SocietySocial Issues & Advocacy
Racial diversity does not disadvantage qualified Asian American applicants
Key takeaways:
The speaker expresses disdain for someone who wants to destroy affirmative action.
There is speculation about whether the person in question donated to Jesse Helms’s Senate race in 1984.
Jesse Helms used a TV ad in his campaign against Harvey Gantt that criticized affirmative action.
The speaker expresses satisfaction at Jesse Helms’s death before returning to the main case.
SFFA argues that racial diversity disadvantage qualified Asian American applicants, which the speaker disagrees with.
Asian Americans make up a significant percentage of Harvard admissions compared to their population size.
In 2008, Asian Americans accounted for a high percentage of top-scoring students admitted to Harvard.
Transcript: Speaker 2 You imagine being a person whose entire adult life like every fantasy you’ve ever had is that you get to be the guy who destroys affirmative action. Like it’s so gross like get a life. Yeah.
Speaker 1 I wonder if he donated to Jesse Helms’s Senate race in 1984.
Speaker 2 In which guessing yeah.
Speaker 1 Helms defeated Harvey Gantt with a you know TV ad that said you were going to get that job but they gave it to a minority and it was uh Jesse Helms a piece of shit. So anyway. Well, God, I’m glad he’s dead. All right. Back to this case. SFFA’s argument was that racial diversity disadvantages qualified Asian American applicants. Spoiler alert it does not. But here’s the argument right. So let’s begin with the mathematical facts. Asian Americans are 5.6% of the population and they average between 17 and 23% of Harvard admissions. And now I’m going to quote from paragraph 218 of the complaint. In 2008 Asian Americans made up 46% of domestic Harvard score senders with SAT scores above 2200. A range from which Harvard draws more than half of its students. And in addition Asian Americans made up an even higher percentage of the very top students they accounted for 55% of domestic Harvard score senders with SAT scores above 2300. These data in combination with other publicly available data demonstrate that Asian Americans admitted to Harvard are
Racial diversity does not disadvantage qualified Asian American applicants Key takeaways: The speaker expresses disdain for someone who wants to destroy affirmative action. Show 447 more
not disadvantaged by Harvard’s admissions policies.
18
Quoted Content
Open
Edit and post a Tweet
Add Autotags

Open
Post Tweet
7/20/2023
Child
parents
Supportive
Accommodation
parents
children
traps
children
parent
parent
things
parenting styles
confidence
traps
takeaways
ability
one
things
parents
choice
anxiety
attitude
parenting styles
Speaker
Transcript
message
trap
teeth
Buttercup
way
 Show 19 more
Child
parents
Supportive
Accommodation
parents
children
traps
children
parent
parent
things
parenting styles
confidence
traps
takeaways
ability
one
things