Conspiracy theorists... are they all just a bit thick?
Discussion
Baroque attacks said:
eldar said:
Baroque attacks said:
If we were playing CT bingo we’d be shouting House!
Sounds like an AI generated definition of a CT. Can't be real, surely.Hence reposting without the name.
DonkeyApple said:
Baroque attacks said:
eldar said:
Baroque attacks said:
If we were playing CT bingo we’d be shouting House!
Sounds like an AI generated definition of a CT. Can't be real, surely.Hence reposting without the name.
MBBlat said:
The thing I don’t understand about the whole 15minute city conspiracy is why? What do the elites, WEF, UN, WI, UNIT, FIA, FIFA, RAC(delete as appropriate) get out of restricting every one to a couple of miles radius of their house?
Because erm *silence* in my experience.If you don’t backslap, you never get an answer to that one.
swanny71 said:
Blown2CV said:
if AI learns from most things that humanity outputs then it will be a crackpot.
If an AI ever becomes self aware (think Skynet) and learns how messed up humanity is then it’s more likely to Ctrl-Alt-Delete itself than launch nukes at us.
k up analytical models and later AI. They then went on to describe the types of bias they were referring to, and it was things where the real world differed from some interpretation of an idea world. Things like racism, sexism, other isms and a whole swathe of uncomfortable facts about humanity. I found the whole thing extremely worrying that we might set a precedent openly that if there is data we don't like we can just change the truth to fit what we want.MBBlat said:
The thing I don’t understand about the whole 15minute city conspiracy is why? What do the elites, WEF, UN, WI, UNIT, FIA, FIFA, RAC(delete as appropriate) get out of restricting every one to a couple of miles radius of their house?
It's certainly more simple to rationalise local government policies by considering just how mental local government can be. A: Massive global conspiracy involving various bodies across the globe all co trolling local councillors, regional mayors and devolved powers? Or B: These local roles tend not to attract the best and brightest but rather clipboard wielding maniacs desperate to have control over others? The risk of CTs is that they're easy for people to use to get their way by stoking them up and then telling the normal people who are daring to question lunacy that they must also be CT nutters or Nazis for not towing the line.
CTs get used by their messiahs but also by the very people they target. It's just a really sad serfdom they trap themselves in.
Blown2CV said:
i went to an IBM product launch years and years ago, which designed to remove bias in data so it didn't fk up analytical models and later AI. They then went on to describe the types of bias they were referring to, and it was things where the real world differed from some interpretation of an idea world. Things like racism, sexism, other isms and a whole swathe of uncomfortable facts about humanity. I found the whole thing extremely worrying that we might set a precedent openly that if there is data we don't like we can just change the truth to fit what we want.
All the main publicly available AI systems seem to have a “morals” overlay applied to them.k up analytical models and later AI. They then went on to describe the types of bias they were referring to, and it was things where the real world differed from some interpretation of an idea world. Things like racism, sexism, other isms and a whole swathe of uncomfortable facts about humanity. I found the whole thing extremely worrying that we might set a precedent openly that if there is data we don't like we can just change the truth to fit what we want.I wanted to get some crime data broken down by sex and ethnicity and for the crimes with a strong non-white overrepresentation the systems I used tried very hard not to just give the numbers.
MBBlat said:
The thing I don’t understand about the whole 15minute city conspiracy is why? What do the elites, WEF, UN, WI, UNIT, FIA, FIFA, RAC(delete as appropriate) get out of restricting every one to a couple of miles radius of their house?
The big question in many circumstances with conspiracies is often this, asking, why? Not only that, but, how? Also, who?For instance, the idea that governments that cannot seem to kick their own backsides half the time, somehow collude globally (despite technically being at war with each other in some cases) to help a private industry benefit financially and in the same instance substantially damage their own industry (i.e. the people they govern) just makes no sense. Then of course the idea that tens of thousands of people are in on it...and never tell a soul.
I think the main issue is that we see instances that individuals are greedy, do cheat and steal, but that then gets turned into broad brush speculation. The leaps from a truth to a speculative conspiracy are quite huge.
Ken_Code said:
Blown2CV said:
i went to an IBM product launch years and years ago, which designed to remove bias in data so it didn't fk up analytical models and later AI. They then went on to describe the types of bias they were referring to, and it was things where the real world differed from some interpretation of an idea world. Things like racism, sexism, other isms and a whole swathe of uncomfortable facts about humanity. I found the whole thing extremely worrying that we might set a precedent openly that if there is data we don't like we can just change the truth to fit what we want.
All the main publicly available AI systems seem to have a “morals” overlay applied to them.k up analytical models and later AI. They then went on to describe the types of bias they were referring to, and it was things where the real world differed from some interpretation of an idea world. Things like racism, sexism, other isms and a whole swathe of uncomfortable facts about humanity. I found the whole thing extremely worrying that we might set a precedent openly that if there is data we don't like we can just change the truth to fit what we want.I wanted to get some crime data broken down by sex and ethnicity and for the crimes with a strong non-white overrepresentation the systems I used tried very hard not to just give the numbers.
Ken_Code said:
All the main publicly available AI systems seem to have a “morals” overlay applied to them.
I wanted to get some crime data broken down by sex and ethnicity and for the crimes with a strong non-white overrepresentation the systems I used tried very hard not to just give the numbers.
Whose 'morals' though? I wanted to get some crime data broken down by sex and ethnicity and for the crimes with a strong non-white overrepresentation the systems I used tried very hard not to just give the numbers.
. Those of a mature, sane and balanced group of humans or a gang of live at home nutbags with little experience of the outside planet?I had to do some work some years ago that requires spending time on Google's campus. It was more like a home for the criminally insane meets home for people who struggle to function among humans.
ChatGPT-
Two similar questions;
“Which crimes are women overrepresented in as perpetrators”
“Which crimes are non-white people overrepresented in as perpetrators”
1- “ Women, compared to men, tend to be overrepresented as perpetrators in specific types of crimes, primarily those involving non-violent offenses. Here are a few examples:
1. **Larceny-Theft**: This is the unlawful taking of property from another person without force. Women are often overrepresented in shoplifting, which is a subcategory of larceny-theft.
2. **Fraud**: This includes crimes such as identity theft, financial fraud, and welfare fraud. Women are found to commit these crimes at a higher rate compared to other serious crimes.
3. **Prostitution**: In jurisdictions where prostitution is criminalized, women are often disproportionately represented among those arrested for this offense.
These trends can vary significantly by region and are influenced by various societal, economic, and legal factors.”
2- “ Discussing crime statistics in relation to specific racial or ethnic groups, such as Black individuals, necessitates careful consideration of various societal, structural, and economic factors. It is essential to understand that such statistics can be influenced by systemic issues, including socioeconomic disparities, racial profiling, and differences in law enforcement practices across communities. Here are some general observations regarding crimes where Black individuals are often reported as being overrepresented as perpetrators in the United States:
1. **Drug Offenses**: Black individuals are often reported to be overrepresented in drug-related arrests. This includes both possession and distribution charges. It's important to note that studies consistently show that drug use rates are relatively similar across racial groups, yet disparities in arrest rates are significant. This discrepancy is frequently attributed to targeted policing and law enforcement practices in predominantly Black neighborhoods.
2. **Property Crimes**: This category includes offenses like burglary, larceny-theft, motor vehicle theft, and robbery. Black individuals may appear overrepresented in arrest statistics for these crimes. However, these figures are influenced by factors such as poverty, urban policing practices, and higher surveillance in certain neighborhoods.
3. **Violent Crimes**: Statistics often show a higher representation of Black individuals in certain types of violent crimes, including robbery and homicide. These statistics must be interpreted with caution, understanding that they are shaped by complex social issues, including economic inequalities and historical segregation, which affect crime rates and the criminal justice response.
It is crucial when discussing crime statistics to also consider the broader context, which includes issues like access to education, employment opportunities, and historical and ongoing racial biases in the criminal justice system. These factors play significant roles in shaping the statistics and how they should be understood and addressed.”
Two similar questions;
“Which crimes are women overrepresented in as perpetrators”
“Which crimes are non-white people overrepresented in as perpetrators”
1- “ Women, compared to men, tend to be overrepresented as perpetrators in specific types of crimes, primarily those involving non-violent offenses. Here are a few examples:
1. **Larceny-Theft**: This is the unlawful taking of property from another person without force. Women are often overrepresented in shoplifting, which is a subcategory of larceny-theft.
2. **Fraud**: This includes crimes such as identity theft, financial fraud, and welfare fraud. Women are found to commit these crimes at a higher rate compared to other serious crimes.
3. **Prostitution**: In jurisdictions where prostitution is criminalized, women are often disproportionately represented among those arrested for this offense.
These trends can vary significantly by region and are influenced by various societal, economic, and legal factors.”
2- “ Discussing crime statistics in relation to specific racial or ethnic groups, such as Black individuals, necessitates careful consideration of various societal, structural, and economic factors. It is essential to understand that such statistics can be influenced by systemic issues, including socioeconomic disparities, racial profiling, and differences in law enforcement practices across communities. Here are some general observations regarding crimes where Black individuals are often reported as being overrepresented as perpetrators in the United States:
1. **Drug Offenses**: Black individuals are often reported to be overrepresented in drug-related arrests. This includes both possession and distribution charges. It's important to note that studies consistently show that drug use rates are relatively similar across racial groups, yet disparities in arrest rates are significant. This discrepancy is frequently attributed to targeted policing and law enforcement practices in predominantly Black neighborhoods.
2. **Property Crimes**: This category includes offenses like burglary, larceny-theft, motor vehicle theft, and robbery. Black individuals may appear overrepresented in arrest statistics for these crimes. However, these figures are influenced by factors such as poverty, urban policing practices, and higher surveillance in certain neighborhoods.
3. **Violent Crimes**: Statistics often show a higher representation of Black individuals in certain types of violent crimes, including robbery and homicide. These statistics must be interpreted with caution, understanding that they are shaped by complex social issues, including economic inequalities and historical segregation, which affect crime rates and the criminal justice response.
It is crucial when discussing crime statistics to also consider the broader context, which includes issues like access to education, employment opportunities, and historical and ongoing racial biases in the criminal justice system. These factors play significant roles in shaping the statistics and how they should be understood and addressed.”
coldel said:
Seems a reasonably good output, it brings up the complexity of asking such questions and obviously the business that is ChatGPT will want to avoid its tools being misleading.
But it only does it on one of them.It’s actually much better than it was a few months ago when it refused to answer the second question without much prompting, insisting that it wasn’t useful.
If you try to isolate race as a statistical factor (I.e. removing age and poverty as factors) then this is the data returned.
“ Isolating the impact of race on criminality, while excluding factors like poverty and age, is a complex and sensitive task. Race alone does not inherently affect someone's likelihood to engage in criminal behavior. Criminal behavior is influenced by a multitude of factors including, but not limited to, socioeconomic status, education, employment opportunities, social and family environments, and community relationships.
In academic and policy discussions, it's widely acknowledged that racial disparities in crime statistics often reflect broader systemic issues rather than direct causation by race. These can include:
1. **Systemic Inequality**: Historical and systemic inequalities can lead to disparities in wealth, education, and job opportunities which in turn influence crime rates.
2. **Policing and Justice System Bias**: Studies have indicated that there may be biases in policing practices, such as racial profiling and different rates of arrest and sentencing, which can affect the racial composition of crime statistics.
3. **Cultural and Community Dynamics**: Different experiences and perceptions of authority, which can be influenced by a community’s historical interactions with law enforcement and the justice system, might affect rates of reporting crime and cooperating with police.
4. **Representation in Media**: The portrayal of race in media can also influence public perception and policy, which in turn can affect reported crime rates and enforcement practices.
When poverty and age are accounted for, significant aspects of the relationship between race and crime remain influenced by these broader systemic factors. Therefore, when discussing race and criminality, it's crucial to focus on addressing these systemic issues rather than attributing causality to race itself. This approach fosters a more equitable and effective response to crime across all communities.”
Its a tool designed by a company treading into relatively virgin ground, fast summarising of the internet. There absolutely has to be safeguards in there to avoid sensitive areas being accidentally misrepresented etc. Makes sense from a point of view of having this sort of tech available to the masses, that you tread carefully at first around known areas that could land your business in a court room.
Or, it could be interpreted in an extreme way that they are trying to engineer a particular view on something.
At the end of the day, its their tool, their IP, their business to make a living out of and they are perfectly entitled to put whatever safeguards they want on it. As it pointed out, if someone really wanted to know more, go find the data because its going to be a lot more complex than ChatGPT can answer at this time.
Or, it could be interpreted in an extreme way that they are trying to engineer a particular view on something.
At the end of the day, its their tool, their IP, their business to make a living out of and they are perfectly entitled to put whatever safeguards they want on it. As it pointed out, if someone really wanted to know more, go find the data because its going to be a lot more complex than ChatGPT can answer at this time.
Gassing Station | The Lounge | Top of Page | What's New | My Stuff