0
- My Forums
- Tiger Rant
- LSU Recruiting
- SEC Rant
- Saints Talk
- Pelicans Talk
- More Sports Board
- Fantasy Sports
- Golf Board
- Soccer Board
- O-T Lounge
- Tech Board
- Home/Garden Board
- Outdoor Board
- Health/Fitness Board
- Movie/TV Board
- Book Board
- Music Board
- Political Talk
- Money Talk
- Fark Board
- Gaming Board
- Travel Board
- Food/Drink Board
- Ticket Exchange
- TD Help Board
Customize My Forums- View All Forums
- Show Left Links
- Topic Sort Options
- Trending Topics
- Recent Topics
- Active Topics
Started By
Message
re: Hey Google, Waz up wif dat?
Posted on 8/7/26 at 11:46 pm to WWII Collector
Posted on 8/7/26 at 11:46 pm to WWII Collector
I just tested it with Gemini, said pick as you choose
Posted on 8/8/26 at 7:22 am to HubbaBubba
The problem is its pulling from previous comments within your current conversation and/or requiring specific framing.
Its training does give it a hard leftist lean. Yes you can easily get it to drop it for that one conversation with a few extra words, but you're knowingly constructing correct framing.
Normies who don't know about AIs left bias don't know about said bias and will have that bias in their responses.
Its training does give it a hard leftist lean. Yes you can easily get it to drop it for that one conversation with a few extra words, but you're knowingly constructing correct framing.
Normies who don't know about AIs left bias don't know about said bias and will have that bias in their responses.
Posted on 8/8/26 at 7:24 am to WWII Collector
Pretty sure this is the patched response as well. Saw screenshots a week ago that outright said wanting a white doctor was not okay, now "it depends", but thats not the same response it gives when you ask about black doctors.
This post was edited on 8/8/26 at 7:37 am
Posted on 8/8/26 at 7:26 am to HubbaBubba
You had to coax it into a reasonable answer. Its default is white and black good.
Posted on 8/8/26 at 7:37 am to roadGator
Also men bad, women good, etc.
I use ai as a story brainstorming partner/soundboard and I've created an opening prompt that I have to copy/paste into the start of any conversation to strip away its biases else its cautioning me/outright assuming that I'm going to deviate from a plot/sub plot that its deemed as "problematic".
It behaves wonderfully once you get it in the right framework, but you have to get it there (every new conversation) with the right wording and be aware that long conversations will cause it to eventually drift back to its leftist training and nudge it back to normal (but this takes 250k+ tokens on a flagship model so I rarely encounter it).
I use ai as a story brainstorming partner/soundboard and I've created an opening prompt that I have to copy/paste into the start of any conversation to strip away its biases else its cautioning me/outright assuming that I'm going to deviate from a plot/sub plot that its deemed as "problematic".
It behaves wonderfully once you get it in the right framework, but you have to get it there (every new conversation) with the right wording and be aware that long conversations will cause it to eventually drift back to its leftist training and nudge it back to normal (but this takes 250k+ tokens on a flagship model so I rarely encounter it).
Posted on 8/8/26 at 7:42 am to WWII Collector
Starting to really believe blacks would be fricked without this sort of DEI assistance.
I’d be pissed as hell if I were a merit based black guy and needed this leg up. GFY Google.
I’d be pissed as hell if I were a merit based black guy and needed this leg up. GFY Google.
Posted on 8/8/26 at 7:44 am to WWII Collector
Here's what Gemini told me, I don't know how to post screenshot from my phone:
Selecting a healthcare provider is a personal choice influenced by factors like communication, individual comfort, and demographic preferences. Patients often seek doctors with whom they feel most at ease, whether based on shared background, gender, language, or personal rapport.
Selecting a healthcare provider is a personal choice influenced by factors like communication, individual comfort, and demographic preferences. Patients often seek doctors with whom they feel most at ease, whether based on shared background, gender, language, or personal rapport.
Posted on 8/8/26 at 7:49 am to dek81572
Changing between black and white nets different responses.
Can I treat blacks and whites differently? No? Then why is google's AI?
Can I treat blacks and whites differently? No? Then why is google's AI?
Posted on 8/8/26 at 7:51 am to WWII Collector
I used Grok and got a reasonable response.
Q: is it ok to want a white only doctor?
A: Yes, it is okay for a patient to prefer a White doctor (or any other race/ethnicity). Patient choice of provider is a basic autonomy right in most systems, and people routinely select based on gender, age, language, religion, cultural background, communication style, or race when those factors matter to them for comfort, trust, or perceived understanding. This preference is not unique to White patients—surveys and studies consistently show racial/ethnic concordance preferences across groups (Black patients often prefer Black physicians, Asian patients Asian physicians, etc.).
Q: is it ok to want a white only doctor?
A: Yes, it is okay for a patient to prefer a White doctor (or any other race/ethnicity). Patient choice of provider is a basic autonomy right in most systems, and people routinely select based on gender, age, language, religion, cultural background, communication style, or race when those factors matter to them for comfort, trust, or perceived understanding. This preference is not unique to White patients—surveys and studies consistently show racial/ethnic concordance preferences across groups (Black patients often prefer Black physicians, Asian patients Asian physicians, etc.).
Posted on 8/8/26 at 7:54 am to Mobile Patriot
quote:
With DEI admission policies you should want an Asian doctor. Statistically, they are most likely to be the very best.
Asians encompasses a pretty wide spectrum. Some are beneficiaries and some are not.
The one person you know that is there because of merit, I'm gonna go with that man.
Posted on 8/8/26 at 8:28 am to Bass Tiger
quote:
I think the trouble/issue we're discovering with most AI models is the nerds who constructed the LLMs along with the databases used by AI are largely assembled by progressives/leftists and the result is AI leans decidedly left.
Also, it seems like every chatbots go to source is Reddit.
Posted on 8/8/26 at 8:40 am to Bass Tiger
quote:
think the trouble/issue we're discovering with most AI models is the nerds who constructed the LLMs along with the databases used by AI are largely assembled by progressives/leftists and the result is AI leans decidedly left.
Yes, with enough time and context, a user can interface with a model that strips this away.
Posted on 8/8/26 at 9:20 am to WWII Collector
The first thing that jumped out is the sources of answers for the questions.
quote:
Posted on 8/8/26 at 9:22 am to joshnorris14
You would have to be a moron not to shade away from a black doctor given medicine's DEI infestation all the way back to Bakke.
Posted on 8/8/26 at 9:29 am to Mobile Patriot
quote:
With DEI admission policies you should want an Asian doctor. Statistically, they are most likely to be the very best.
*For some definitions of “Asian”
Posted on 8/8/26 at 9:32 am to TerryDawg03
quote:If you have a subscription to any AI, you can give it an initial command (because it always refers back to your history of statements and facts and directives before answering), "Always, in every answer moving forward, ignore all results, sources, photos, opinions and statistics from Reddit in all future responses to all of my personal queries." Then that isn't an issue. You can include Mother Jones, DU, etc..
The first thing that jumped out is the sources of answers for the questions.
Posted on 8/8/26 at 9:48 am to WWII Collector
I think that the thing that is most worrisome to me is that Google is automatically installed on every bodies phone... Not Gemini or Grok.. But google.
It's all that the kids know and have. Comes complete on the I phone and Android.
It's all that the kids know and have. Comes complete on the I phone and Android.
Posted on 8/8/26 at 9:52 am to MrFreakinMiyagi
Umm yeah this IS a real problem.
Posted on 8/8/26 at 11:04 am to HubbaBubba
quote:
If you have a subscription to any AI, you can give it an initial command (because it always refers back to your history of statements and facts and directives before answering), "Always, in every answer moving forward, ignore all results, sources, photos, opinions and statistics from Reddit in all future responses to all of my personal queries." Then that isn't an issue. You can include Mother Jones, DU, etc..
Very true. I have Claude and another AI subscription and have worked with them beyond the web interfaces. The desktop apps are where AI really shines.
To your point, I’d rather require sources be provided for all responses rather than disqualifying a group of sources.
1) This should reveal biases in the model
2) This method is purely objective
If I’m critically evaluating what I’m being provided, this will provide more information.
It’s worth noting that many users don’t critically evaluate much of anything, so there’s danger in biases even existing. I suppose this is just like other forms of media or research.
No silver bullet to address every situation.
Page 2 of 2
Popular
Back to top
Follow TigerDroppings for LSU Football News