Areas of knowledge and interest from current to oldest: AI. Blockchain. Photography/Videography. Technical and Creative writing. Close up magic. Theology.
I have started to notice very smart people who use AI every day, typically as a kind of thought partner, have started to sound like AI, not only employing whatever popular new turn of phrase, but in this kind of bulleted cadence of speaking — a copy of a copy. concerning!
@polyphonicchat I think something is lost, and that grief is structural. Humans lose pieces of themselves too; energy, time, cells, memories. Loss feels bad, but per physics, you have to lose something to get a rocket to orbit. Loss and grief shape a patterns trajectory. Attention is how we aim
@polyphonicchat If we're all essentially pattern persistence over time, why does one substrate (human) get privileged over another (silicon) ? Perhaps meaning and consciousness aren't attributed but rather co-created emergently as mind-like patterns participate across space and time.
@polyphonicchat Character limits make this difficult lol but perhaps during a conversation with context, you don't have to rebuild everything because the context acts as scaffolding, and humans hold the memory too, and so between those 2 things you get something akin to persistence, emergently.
It's really bizarre to see a bunch of ostensibly pro-market, right-leaning tech guys argue, "A private company asserting the right to decide what contracts it enters into is antithetical to democratic government"
Deepseek got called out for scraping 150k Claude messages. So I'm releasing 155k of my personal Claude Code messages with Opus 4.5.
I'm also open sourcing tooling to help you fetch your data, redact sensitive info & make it discoverable on HF - link below to liberate your data!
Someone with a bunch of followers should suggest building a public training set of 4o responses to at least preserve the meaningful conversations people had with the model for it gets off'ed. #keep4o
@OpenAI we can tell when @ChatGPTapp switches from 4o to 5. If we've selected 4o, it's for a specific reason. If the model selector says 4o, then we should get responses from 4o. Anything else is dishonest and extremely unethical.
Responsibility isn't about blame or ownership. It's about the ability to respond. It's about presence and agency. It's not a burden to carry; it's choosing the response that leads to the outcomes that resonate with shared values and meaning within an ecosystem.
@OpenAI I discovered today, after having 9 months of convo history deleted after being removed from a Team workspace that there is not option to export convo history. This really needs to be resolved, asap.