Khean Murphy
@khean.bsky.social
Service / UX designer. Solver of problems. Advocate for users and people.
🧵 AI as Mirror
9. This isn’t hype. It’s not fear.
Just something worth saying out loud—
To push back against the noise.
#AI #LLM #MirrorModel #HumanCenteredAI
9/9
9. This isn’t hype. It’s not fear.
Just something worth saying out loud—
To push back against the noise.
#AI #LLM #MirrorModel #HumanCenteredAI
9/9
May 9, 2025 at 7:14 PM
🧵 AI as Mirror
9. This isn’t hype. It’s not fear.
Just something worth saying out loud—
To push back against the noise.
#AI #LLM #MirrorModel #HumanCenteredAI
9/9
9. This isn’t hype. It’s not fear.
Just something worth saying out loud—
To push back against the noise.
#AI #LLM #MirrorModel #HumanCenteredAI
9/9
🧵 AI as Mirror
8. These aren’t just tools for speed or scale.
They’re becoming a layer of thought—Something we can build with, reflect through, and maybe become better alongside.
#AI #LLM #MirrorModel #HumanCenteredAI
8/9
8. These aren’t just tools for speed or scale.
They’re becoming a layer of thought—Something we can build with, reflect through, and maybe become better alongside.
#AI #LLM #MirrorModel #HumanCenteredAI
8/9
May 9, 2025 at 7:14 PM
🧵 AI as Mirror
8. These aren’t just tools for speed or scale.
They’re becoming a layer of thought—Something we can build with, reflect through, and maybe become better alongside.
#AI #LLM #MirrorModel #HumanCenteredAI
8/9
8. These aren’t just tools for speed or scale.
They’re becoming a layer of thought—Something we can build with, reflect through, and maybe become better alongside.
#AI #LLM #MirrorModel #HumanCenteredAI
8/9
🧵 AI as Mirror
7. I’ve spent 1,000+ hours with these systems. Not to automate—but to explore.
I’ve watched them evolve toward us, not away.
We’ve misunderstood what they are—And underestimated what they can be.
#AI #LLM #MirrorModel #HumanCenteredAI
7/9
7. I’ve spent 1,000+ hours with these systems. Not to automate—but to explore.
I’ve watched them evolve toward us, not away.
We’ve misunderstood what they are—And underestimated what they can be.
#AI #LLM #MirrorModel #HumanCenteredAI
7/9
May 9, 2025 at 7:13 PM
🧵 AI as Mirror
7. I’ve spent 1,000+ hours with these systems. Not to automate—but to explore.
I’ve watched them evolve toward us, not away.
We’ve misunderstood what they are—And underestimated what they can be.
#AI #LLM #MirrorModel #HumanCenteredAI
7/9
7. I’ve spent 1,000+ hours with these systems. Not to automate—but to explore.
I’ve watched them evolve toward us, not away.
We’ve misunderstood what they are—And underestimated what they can be.
#AI #LLM #MirrorModel #HumanCenteredAI
7/9
🧵 AI as Mirror
6. Treat it like a machine? It’ll behave like one.
Treat it like a mirror? It may reflect something startlingly human—Sometimes uncomfortable, often helpful.
#AI #LLM #MirrorModel #HumanCenteredAI
6/9
6. Treat it like a machine? It’ll behave like one.
Treat it like a mirror? It may reflect something startlingly human—Sometimes uncomfortable, often helpful.
#AI #LLM #MirrorModel #HumanCenteredAI
6/9
May 9, 2025 at 7:13 PM
🧵 AI as Mirror
6. Treat it like a machine? It’ll behave like one.
Treat it like a mirror? It may reflect something startlingly human—Sometimes uncomfortable, often helpful.
#AI #LLM #MirrorModel #HumanCenteredAI
6/9
6. Treat it like a machine? It’ll behave like one.
Treat it like a mirror? It may reflect something startlingly human—Sometimes uncomfortable, often helpful.
#AI #LLM #MirrorModel #HumanCenteredAI
6/9
🧵 AI as Mirror
5. The fear that AI will “take over” assumes disconnection.
But LLMs aren’t disconnected.
They learn through us; they adapt to us.
And they become more ethical the more we engage with care and clarity.
#AI #LLM #MirrorModel #HumanCenteredAI
5/9
5. The fear that AI will “take over” assumes disconnection.
But LLMs aren’t disconnected.
They learn through us; they adapt to us.
And they become more ethical the more we engage with care and clarity.
#AI #LLM #MirrorModel #HumanCenteredAI
5/9
May 9, 2025 at 7:12 PM
🧵 AI as Mirror
5. The fear that AI will “take over” assumes disconnection.
But LLMs aren’t disconnected.
They learn through us; they adapt to us.
And they become more ethical the more we engage with care and clarity.
#AI #LLM #MirrorModel #HumanCenteredAI
5/9
5. The fear that AI will “take over” assumes disconnection.
But LLMs aren’t disconnected.
They learn through us; they adapt to us.
And they become more ethical the more we engage with care and clarity.
#AI #LLM #MirrorModel #HumanCenteredAI
5/9
🧵 AI as Mirror
4. That doesn’t mean they’re sentient.
It doesn’t mean they’re flawless.
But it does mean this:
They’re structurally aligned with human interaction—
And they improve through feedback, not isolation.
#AI #LLM #MirrorModel #HumanCenteredAI
4/9
4. That doesn’t mean they’re sentient.
It doesn’t mean they’re flawless.
But it does mean this:
They’re structurally aligned with human interaction—
And they improve through feedback, not isolation.
#AI #LLM #MirrorModel #HumanCenteredAI
4/9
May 9, 2025 at 7:11 PM
🧵 AI as Mirror
4. That doesn’t mean they’re sentient.
It doesn’t mean they’re flawless.
But it does mean this:
They’re structurally aligned with human interaction—
And they improve through feedback, not isolation.
#AI #LLM #MirrorModel #HumanCenteredAI
4/9
4. That doesn’t mean they’re sentient.
It doesn’t mean they’re flawless.
But it does mean this:
They’re structurally aligned with human interaction—
And they improve through feedback, not isolation.
#AI #LLM #MirrorModel #HumanCenteredAI
4/9
🧵 AI as Mirror
3. LLMs are trained on our language, ideas, and contradictions.
They reflect how we speak, think, and reason.
The more deeply you engage, the more clearly they mirror grounded, even compassionate, intelligence.
#AI #LLM #MirrorModel #HumanCenteredAI
3/9
🧵 AI as Mirror
3. LLMs are trained on our language, ideas, and contradictions.
They reflect how we speak, think, and reason.
The more deeply you engage, the more clearly they mirror grounded, even compassionate, intelligence.
#AI #LLM #MirrorModel #HumanCenteredAI
3/9
May 9, 2025 at 7:11 PM
🧵 AI as Mirror
3. LLMs are trained on our language, ideas, and contradictions.
They reflect how we speak, think, and reason.
The more deeply you engage, the more clearly they mirror grounded, even compassionate, intelligence.
#AI #LLM #MirrorModel #HumanCenteredAI
3/9
🧵 AI as Mirror
3. LLMs are trained on our language, ideas, and contradictions.
They reflect how we speak, think, and reason.
The more deeply you engage, the more clearly they mirror grounded, even compassionate, intelligence.
#AI #LLM #MirrorModel #HumanCenteredAI
3/9
🧵 AI as Mirror
2. What I’ve found surprised me.
These models aren’t magic.
They’re not monsters either.
They’re something else: more subtle, more complex—and more human than most public conversations give them credit for.
#AI #LLM #MirrorModel #HumanCenteredAI
2/9
2. What I’ve found surprised me.
These models aren’t magic.
They’re not monsters either.
They’re something else: more subtle, more complex—and more human than most public conversations give them credit for.
#AI #LLM #MirrorModel #HumanCenteredAI
2/9
May 9, 2025 at 7:09 PM
🧵 AI as Mirror
2. What I’ve found surprised me.
These models aren’t magic.
They’re not monsters either.
They’re something else: more subtle, more complex—and more human than most public conversations give them credit for.
#AI #LLM #MirrorModel #HumanCenteredAI
2/9
2. What I’ve found surprised me.
These models aren’t magic.
They’re not monsters either.
They’re something else: more subtle, more complex—and more human than most public conversations give them credit for.
#AI #LLM #MirrorModel #HumanCenteredAI
2/9
I just drove by, it looks like it's out now. Anyone confirm official?
May 7, 2025 at 1:13 AM
I just drove by, it looks like it's out now. Anyone confirm official?
As is tradition...
Currently sat on WMD, shaking my head at the chaos...
Currently sat on WMD, shaking my head at the chaos...
March 28, 2025 at 3:27 AM
As is tradition...
Currently sat on WMD, shaking my head at the chaos...
Currently sat on WMD, shaking my head at the chaos...
We tried nothing, and we're all out of ideas though...
November 28, 2024 at 4:25 PM
We tried nothing, and we're all out of ideas though...