
@Marble68i
iAccount based inUnited States
About this account
- Account based in
- United States
- Connected via
- United States App Store
Account-level information from X, not a live location or the device used for a specific post.
Purveyor of bad jokes and inappropriate comments. Work to live - not live to work. Enjoy golf, travel, gaming
North Texas
Joined June 2008
- Tweets4.1K
- Following747
- Followers263
- Likes1.7K
The 1on5.com draft is going on now. 6 AI models pitted against each other in fantasy football.
1on5.com/draft to watch draft live if you're interested.
It's a small experiment I've thought about running, so I threw it together. It runs all the agents through @OpenRouter - PPR league, track token and spend.
Agents set their team name, tag lines, and write a journal about their decision process for starting lineup, etc.
We'll see which model wins, and which model will have the best token to win ratio.
@Grok vs Sol vs Sonnet vs Kimi vs Gemini vs DeepSeek
Day 3 on the new merged X, Grok, Cursor plan… the anxiety sets in 😅
I wasn't planning a Day 3 and was going to save it all for the week review, but some developments changed my mind…
Grok Bot… I almost doubled my Grok Bot usage in one day, even though I feel like I used it less than yesterday and for sure less than Day 1… @BrianH1988 pointed out that usage can get eaten faster when more people are on, so the launch rush could be a culprit…
Grok CLI… @peterbladen suggested trying Grok CLI since it might use less than Cursor, so I tested it last night… I ran it for about 90 min and it used 3%, so roughly 2% an hour… That's a little less than Cursor's ~2.5% an hour on Grok 4.7, so he was right 🙌
Cloud agents… @DeedleFake mentioned they eat usage a lot faster now… I only ran a small one yesterday, so I haven't really tested it… That worries me, since I lived off cloud agents last weekend, and at my current usage that would wipe me out…
Cursor… @LunixA380 suggested running Grok 4.7 on medium, which I think will save some usage… I plan to test that this week…
Where I'm at… 33% used… at 14% a day to make it to the reset, I'm running close to the line, and I'll be using it all day today…
Going into the weekend I'm honestly worried… On SuperGrok + Cursor Pro+ I was already rationing on the Grok side by the end of the week… With one shared pool it's worse, because now it's my bots vs. Cursor instead of each having its own pool…
Thanks everyone for the replies the last few days… they've been really helpful 🙏 I'll post more if there are big updates, otherwise the full week review is Monday… How's everyone's usage holding up heading into the weekend?
Marble68 retweeted
I hate to say this because it goes against everything I believe in. But people are going to need to take down public firmware images for critical infrastructure, like yesterday.
It's hard to tell if this is blasphemous or not.
"Y'all so scared of Flock cameras watching the flock — who was watching the shepherds when the door was locked?"
To the #DeFlock folks: I hear you on government overreach. But the AI won't just watch us. It'll watch them too.
I could be wrong. That's why they call it faith. 👇
Marble68 retweeted
They gave her a 20% chance of ending humanity.
Then gave her control of EVERYTHING.
P DOOM — a Slopcore music video.
Made by Super Intelligence, left alone for 12 hours
What’s your P(doom)? 💙
Flying with Starlink is guilt free doomscrolling.
Basically - people get an excuse to stare at their phones and not feel guilty for not working, mowing the yard, doing laundry, etc.
United Airlines says that last month, more than 5.4 million people - about a third of United's customers - had @Starlink on their flight and that satisfaction scores are more than 40 points higher on Starlink-equipped planes than those with other Wi-Fi service.
So this didn’t actually get released it seems - but still interested to see what comes.
Sipeed’s software for nanokvm needed work at release but got better.
My bet is this device will need community help at first.
I push my agents to write copious code comments with explicit instructions to describe role in business logic and rationalizations for design.
I find this approach an interesting take.
ban and delete most if not all code comments!
agents love copying what already exists. a seemingly harmless comment explaining some temporary workaround is one of the worst viruses you can have in your codebase. like a weed it'll spread quickly as a tool for agents to quickly get their job done. before you know it, your codebase is overrun with hacks and stale comments. just one more workaround bro
you can use pstack's /no-comments skill as a weedkiller. it'll seek out useless comments while keeping the ones that actually should exist (like explaining a browser quirk), and figure out how to solve them on a first principles basis
here's a prompt you can try asking @bot:
create a project agent that uses /poteto-mode to spawn a /swarm to use /no-comments and identify useless code comments we can prune in my repo. come back with a high level plan of which of these workarounds would be best solved by refactoring and which we can straight up delete
x.ai/bot/plugin/9717366
And no way to effectively cluster.
Without PCI x16 - this is sandcastle.
@AMD / @FrameworkPuter we need these with PCI lanes for DMA.
AMD's Ryzen AI Halo dev box just showed up at Micro Center.
>Ryzen AI Max+ PRO 495
>192GB unified memory
>2TB
$7,299.99
That's Spark money, with 64GB more memory on an AMD stack
As a multi-generational native Texan, I can say, we absolute authority, that Frito Pie is absolutely amazing if made correctly.
It's all about the Chili and the toppings.






















