i'm one of these AI homelab guys but i've been doing it backwards. years of building with agents and explaining it to normies at the farmers market. meanwhile online i'm trading slop memes to fit in because its the only small talk i know how to make. time to fix that.
Okay so there may be a real story here...
@loktar00 covered ~2 weeks ago a sudden and very unusual surge in buying activity on the DGX Sparks at Micro Center. We couldn’t make sense of it at the time because it was unlike anything we’d ever seen.
The entire national supply went to zero in a few days and the surge seemed to come from out of nowhere.
Now some changes are coming to light. Nvidia has officially repriced the DGX Spark to $6,950 MSRP on its website and just today announced a new 64GB DGX Spark for $5,000 with half the memory and a higher price than the 128GB version two weeks ago.
Is it possible that the largest surge in buying that we’ve ever seen just so happened to front run these changes and announcements by two weeks?
Possibly.
But the chance that a group of people knew something early feels like a very real possibility.
https://t.co/0xInkcFQRl
https://t.co/MTKz9bbX0o
Wow - I guess what I called an influencer hyped run on existing stock was actually prudent purchasing and me not believing the current realities of the hardware market. I’ll be eating my hat for lunch today.
Congrats to early DGX Spark owners. Same 128GB box was $3,999 at launch. Founders Edition is now $6,950, and the new 64GB starts at $4,999 with half the RAM.
https://t.co/Fi1lekWWbw
Anyone have the sense we witnessed an influencer manufactured run on dgx sparks happen here over the past couple of days? y’all are really bought into a perpetual hardware inflation scheme? Seems like a blip in the timeline and cheap/free streamed inference remains bountiful. 🤷
@signulll This seems spot on to me. Back in my v1 Claude Code psychosis days last winter I had toyed with a programmatic/AI driven “adaptive surface” layer to replace the surely soon-to-be extinct UI layer. Soon.
@kanishkrazdan Very much the feeling I’ve had looking at this tool side by side with Hermes. sure, it looks like the tool I can hand my Luddite pals, but knowing (or actually, not knowing) what’s happening under the hood makes me take a serious pause.
i used instinct pretty aggressively on day one. it felt magical. and then it turned scary.
giving it access to my email was already scary. then it started establishing sessions across apps and websites i use, on its own.
watching an ai get that much access to act as me made me appreciate granular permissions on APIs a lot more.
think about it. would you give even a clone of yourself unrestricted access to tools that could cost you money & happiness?
i'd feel more comfortable with a dedicated app where sessions stay on device. but even then, i'd want it to ask before logging into anything on my behalf.
and honestly, whatsapp/imessage doesn't feel like the right interface for this.
i want to see where it's working, what it's doing in real time, and be able to step in before it does something i wouldn't have.
that's why i prefer codex for example. the form factor is built around you trying to code, lives in projects and connects to your IDE easily. i'm not completely in the dark about what it's doing.
ai agents don't have to live inside the apps or interfaces we already use.
familiar interfaces help with adoption. they don't solve trust or control.
if something can act as me, i need more than a chat bubble telling me it's done.
i've been running hermes since early and honestly don't use it a ton for stuff like this. lately i use glm 5.3 flash for managing my vps where hermes lives and i cook self-hosted experiments.