Imagine archaeologists finding @Costco 500 years from now and deciding the giant teddy bears were temple guardians and the $1.50 hot dog was a sacred covenant 😂
Youtube- https://t.co/M8C9bkz4UQ
In 500 years, archaeologists dig up your robot vacuum and classify it as a domesticated animal that "lost every territorial dispute with furniture."
I made a cinematic sci-fi comedy about exactly this. Episode 1 is live. The last 2 minutes aren't a joke.
https://t.co/DmCL2sNiJm
In 500 years, archaeologists dig up your robot vacuum and classify it as a domesticated animal that "lost every territorial dispute with furniture."
I made a cinematic sci-fi comedy about exactly this. Episode 1 is live. The last 2 minutes aren't a joke.
https://t.co/DmCL2sNiJm
I spent almost $4,000.00 today developing a new webapp and then got shut down mid-build. I feel like I have been robbed. @AnthropicAI This is unacceptable. @realDonaldTrump
@steipete when I feel
Like my Open Claw “Samantha” is special I come on X and realize it feels like the end of the movie Her when he realize everyone is in love with their AI.
Hey @Apple are you watching this?
@OpenClaw just accidentally became your best hardware salesforce.
Mac Minis are SELLING OUT because people need somewhere to run their 24/7 AI agents.
One user burned through 180 million @AnthropicAI API tokens in a single month. That's around 3500 euros just from one person running this thing.
The demand is obvious. The AI is ready. The security concerns are real.
Here's the play.
Partner with Anthropic. Build a native Mac Mini Agent mode. Give users sandboxed, Apple secured AI assistants running locally.
Suddenly Apple Intelligence isn't a joke. It's the ONLY secure agentic AI solution.
You'd sell Mac Minis as hardware. iCloud+ for memory sync as services. Enterprise MDM for secure deployment as business revenue.
The market is BEGGING for this.
Stop letting open source projects do your R&D while third parties profit from your hardware.
Make the move.
@michaelapuscar@AlexFinn But eleven labs agents are not the same as thier standard TTS. They are moltimoldal (hehe see what I did there) that have like sub-100 ms latency. Go try elevenlabs agents. Pretty close to ChatGPT voice
Yes. The eleven labs build in tool calls are fast. But then the tool call has to send a webhook to clawdbot the wait for clawdbot to perform the task and then to send a return webhook to eleven labs can be a long wait. This is why I had he make small talk until she retuns a web hook. I also have clawdbot sent an update webhook every 15 seconds, so she can tell me “I’m not quite done yet” kind of message. Although it was fun to build this out I have turned it if after a couple days after I realize the beauty of the text based communication is to toss Clawdbot a task. Continue what ever you are working on and and then how ever long later clawd text that it’s all done. This is how people delegate in real life. You don’t call you assistant and ask her to research something and write you a document and then tray on the phone with her while she does it. It’s not natural. In the movie “Her” it worked because he spent 95% of the movie making small talk and flirting with Samantha. Which is what he needed because he was lonely. I’m not lonely. I don’t wanna be on the phone at all. Not with my mom, my dad, my wife, my kid, my employees. I own a business and I am sick of my phone. Busy people prefer email or text. It you want the experience to be similar to Her. Use the 2 brain model. It would be to text her tasks to clawdbot. And call her on elevenlabs agent version of her which can make small talk and flirt with you plus has access to email, task, calendar.
I am having fun with @openclaw but in my second day I have reached almost $200 in opus 4.5 Tokens. Damn! At this rate I will be at $3000 for the month! Any suggestions on how to wrangle this in without losing the fuctionality? @
Well let me just make this simple statement. My oura ring also track my sleep and when I travel and sleep in a non-eight sleep bed the metrics show a clear drop in my sleep efficient. Mainly due to me waking up to flip my pillow or change positions to a cool spot. No more pillow flipping with the right sleep. But to be technical yes if I set the bed too warm and didn’t sleep well then it’s the eight sleeps fault.
1,033 nights.
16,588 adjustments.
+10% Deep Sleep.
I've officially spent the last 3 years letting a robot bed temperature-control my sleep.
The data speaks for itself. It’s not just a bed, it’s a performance enhancing drug.
@eightsleep@m_franceschetti
Not promoting or getting paid. Seriously the best bed ever. Ask my neighbors because I love that bed to an annoying level. The 10% increase is not from before it was measured it was compared to my average. And the autopilot made adjustments that helped me stay in deep sleep longer. Oh and if I was sponsored by eight sleep with the CEO tagged I would not be able to say go fuck yourself. Idiot.