@cakerug.bsky.social
Enjoying asking Gemini questions about technical videos I'm watching on YouTube. But it's missing the share feature so I can save convos as links elsewhere for future reference. Excited for the tools to get better across these LLM chat apps.
February 1, 2025 at 11:44 PM
Enjoying asking Gemini questions about technical videos I'm watching on YouTube. But it's missing the share feature so I can save convos as links elsewhere for future reference. Excited for the tools to get better across these LLM chat apps.
How it has impacted the way I learn: I used to find dangling questions I have immensely distracting from focusing. I could never entertain them before because the tangents I would go on always took too long and were too unwieldy. LLMs let me answer those dangling questions.
For me, LLMs have changed very small things in the way I work, learn, and think. Those small changes have had a massive impact though.
January 31, 2025 at 9:16 PM
How it has impacted the way I learn: I used to find dangling questions I have immensely distracting from focusing. I could never entertain them before because the tangents I would go on always took too long and were too unwieldy. LLMs let me answer those dangling questions.
Some of those changes are simply a byproduct of the chat interface that LLMs are often packaged into.
For me, LLMs have changed very small things in the way I work, learn, and think. Those small changes have had a massive impact though.
January 31, 2025 at 9:08 PM
Some of those changes are simply a byproduct of the chat interface that LLMs are often packaged into.
For me, LLMs have changed very small things in the way I work, learn, and think. Those small changes have had a massive impact though.
January 31, 2025 at 9:06 PM
For me, LLMs have changed very small things in the way I work, learn, and think. Those small changes have had a massive impact though.
Reposted
We grow up with the unacceptable, and accommodation becomes a tool we use to make our lives easier. Accommodating the unacceptable is often easier than doing the right thing. Eventually, we use it to let ourselves off the hook, because that’s easier, too.
www.marytrump.org/p/there-is-n...
www.marytrump.org/p/there-is-n...
There Is No Worst
Opposition in the age of accommodation
www.marytrump.org
January 26, 2025 at 4:58 AM
We grow up with the unacceptable, and accommodation becomes a tool we use to make our lives easier. Accommodating the unacceptable is often easier than doing the right thing. Eventually, we use it to let ourselves off the hook, because that’s easier, too.
www.marytrump.org/p/there-is-n...
www.marytrump.org/p/there-is-n...
Reposted
AI makes sense for incredibly complicated math and medical technology. Writing is how we learn to construct thoughts and ideas. Art is how we learn to express our humanity. AI for the humanités will create a generation that does know how to think or express thoughts.
January 17, 2025 at 2:50 PM
AI makes sense for incredibly complicated math and medical technology. Writing is how we learn to construct thoughts and ideas. Art is how we learn to express our humanity. AI for the humanités will create a generation that does know how to think or express thoughts.
Reposted
The AI madness is so far out of control that "Our product wrote a completely false news alert and credited a real news source" doesn't even rate a corporate comment, let alone an ashamed apology and pulling the product www.bbc.com/news/article...
BBC complains to Apple over misleading shooting headline
Apple's new artificial intelligence features falsely made it seem the BBC reported Luigi Mangione had shot himself.
www.bbc.com
December 13, 2024 at 11:41 PM
The AI madness is so far out of control that "Our product wrote a completely false news alert and credited a real news source" doesn't even rate a corporate comment, let alone an ashamed apology and pulling the product www.bbc.com/news/article...
Reposted
Watching TV news each night and their coverage of the Luigi Mangione case is honestly kind of shocking at this point. I’ve never seen a bigger gap between the way it’s described on TV and the way it’s discussed online and in real life. I always like to think of historical parallels but I’m failing.
December 11, 2024 at 4:29 AM
Watching TV news each night and their coverage of the Luigi Mangione case is honestly kind of shocking at this point. I’ve never seen a bigger gap between the way it’s described on TV and the way it’s discussed online and in real life. I always like to think of historical parallels but I’m failing.
"Evaluation is not the same as thought" CGP Grey hitting the nail on the head about one of the downsides of AI in the latest episode of Cortex.
I Fight Prime Numbers - Cortex
CGP Grey and Myke Hurley are both independent content creators. Each episode, they get together to discuss their working lives. Hosted by CGP Grey and Myke Hurley.
pca.st
December 10, 2024 at 11:45 PM
"Evaluation is not the same as thought" CGP Grey hitting the nail on the head about one of the downsides of AI in the latest episode of Cortex.
Reposted
"Sora is a data-driven physics engine."
x.com/chrisoffner3...
x.com/chrisoffner3...
December 10, 2024 at 12:42 PM
"Sora is a data-driven physics engine."
x.com/chrisoffner3...
x.com/chrisoffner3...
I use LLMs today mostly as a better Google 80% of the time.
It's still very worth it imo.
It's still very worth it imo.
December 8, 2024 at 6:43 PM
I use LLMs today mostly as a better Google 80% of the time.
It's still very worth it imo.
It's still very worth it imo.
Thinking about imposter syndrome.
December 8, 2024 at 6:38 PM
Thinking about imposter syndrome.
Reposted
Great post that captures the tension between classic ML approaches and modern deep learning while acknowledging the nuances of both.
“Working with LLMs doesn’t feel the same. It’s like fitting pieces into a pre-defined puzzle instead of building the puzzle itself.”
www.reddit.com/r/MachineLea...
“Working with LLMs doesn’t feel the same. It’s like fitting pieces into a pre-defined puzzle instead of building the puzzle itself.”
www.reddit.com/r/MachineLea...
From the MachineLearning community on Reddit
Explore this post and more from the MachineLearning community
www.reddit.com
December 6, 2024 at 2:13 AM
Great post that captures the tension between classic ML approaches and modern deep learning while acknowledging the nuances of both.
“Working with LLMs doesn’t feel the same. It’s like fitting pieces into a pre-defined puzzle instead of building the puzzle itself.”
www.reddit.com/r/MachineLea...
“Working with LLMs doesn’t feel the same. It’s like fitting pieces into a pre-defined puzzle instead of building the puzzle itself.”
www.reddit.com/r/MachineLea...
Just found out there's a cap on memory for ChatGPT. Curious to check out other LLMs now since there's less stickiness in ChatGPT for me.
December 4, 2024 at 3:26 PM
Just found out there's a cap on memory for ChatGPT. Curious to check out other LLMs now since there's less stickiness in ChatGPT for me.
I thought I had gotten to the bottom of the reality tv barrel with MILF manor, but now I've found "My husband's not gay". I can't look away.
December 3, 2024 at 1:45 PM
I thought I had gotten to the bottom of the reality tv barrel with MILF manor, but now I've found "My husband's not gay". I can't look away.
Not sure about anti-AI activists but the attack vector of "poisoning" datasets is interesting...
What an interesting twist!
It has occurred to me that anti-AI activists might use PHI and public datasets in order to “poison” indiscriminate data crawling & harvesting by the big labs.
Don’t want to talk about it too much publicly for obvious reasons, but it seems like it could work.
It has occurred to me that anti-AI activists might use PHI and public datasets in order to “poison” indiscriminate data crawling & harvesting by the big labs.
Don’t want to talk about it too much publicly for obvious reasons, but it seems like it could work.
Licensing editorial content for training AI is a violation of the GDPR, according to the Italian watchdog.
"The digital archives of newspapers contain the stories of millions of people, with information, details and even extremely sensitive personal data"
"The digital archives of newspapers contain the stories of millions of people, with information, details and even extremely sensitive personal data"
November 30, 2024 at 2:30 PM
Not sure about anti-AI activists but the attack vector of "poisoning" datasets is interesting...
Reposted
Misogyny is a skill issue.
November 29, 2024 at 11:40 AM
Misogyny is a skill issue.
Everyone knows... What skeet is right?
November 28, 2024 at 2:40 PM
Everyone knows... What skeet is right?
Partway through my first watch through of Downton Abbey and, every once in a while throughout the day, I find myself wondering why tf Anna is so into Mr. Bates.
Currently part way through the prison arc and I'm starting to think he did it.
Currently part way through the prison arc and I'm starting to think he did it.
November 28, 2024 at 1:55 PM
Partway through my first watch through of Downton Abbey and, every once in a while throughout the day, I find myself wondering why tf Anna is so into Mr. Bates.
Currently part way through the prison arc and I'm starting to think he did it.
Currently part way through the prison arc and I'm starting to think he did it.
It's interesting seeing people pile love on Bluesky while at the same time seeing others gathering followers in a classic influencer land grab. It's not all kumbaya, it's still capitalism.
November 28, 2024 at 1:45 PM
It's interesting seeing people pile love on Bluesky while at the same time seeing others gathering followers in a classic influencer land grab. It's not all kumbaya, it's still capitalism.
Reposted
Oh shit, is this how they're testing the capability of the latest models now?? Not who can do it, but who can do it the fastest???
November 28, 2024 at 5:37 AM
Oh shit, is this how they're testing the capability of the latest models now?? Not who can do it, but who can do it the fastest???
I'm already doing this. It's the reason I pay for ChatGPT -- because there is memory and I'm banking on that being wildly useful in the future.
Plus my immigrant mindset is making me feel that this is coming no matter what, let me make sure I know how to use it so it doesn't take my job.
Plus my immigrant mindset is making me feel that this is coming no matter what, let me make sure I know how to use it so it doesn't take my job.
this is the future. People will want their data in because it will be extremely valuable. You can opt out. This is the best possible future
imagining a future state in which people will be clamoring to have their content trained on by AI and that not having your content trained on will be an indicator that it isn’t notable or interesting (not totally dissimilar to how most people want to be ranked in google search)
November 28, 2024 at 12:48 PM
I'm already doing this. It's the reason I pay for ChatGPT -- because there is memory and I'm banking on that being wildly useful in the future.
Plus my immigrant mindset is making me feel that this is coming no matter what, let me make sure I know how to use it so it doesn't take my job.
Plus my immigrant mindset is making me feel that this is coming no matter what, let me make sure I know how to use it so it doesn't take my job.
Widespread topics on my corner of bluesky I've witnessed so far:
- how much we love bluesky
- reading is the best, audiobooks are reading, I don't get reading
- bluesky is an echo chamber but who cares
- AI is evil, AI is good, we need good people on AI or it will be used for evil
- how much we love bluesky
- reading is the best, audiobooks are reading, I don't get reading
- bluesky is an echo chamber but who cares
- AI is evil, AI is good, we need good people on AI or it will be used for evil
November 28, 2024 at 12:36 PM
Widespread topics on my corner of bluesky I've witnessed so far:
- how much we love bluesky
- reading is the best, audiobooks are reading, I don't get reading
- bluesky is an echo chamber but who cares
- AI is evil, AI is good, we need good people on AI or it will be used for evil
- how much we love bluesky
- reading is the best, audiobooks are reading, I don't get reading
- bluesky is an echo chamber but who cares
- AI is evil, AI is good, we need good people on AI or it will be used for evil