have to go through about 8-10 threads like this to start 1 new chat
send 1 message -> pray it works -> repeat in new threads until it finally gets through 1 compaction + response back to me then pray it doesn’t break again next message/compaction (it usually does)
high, high fast, xhigh, xhigh fast. all the same issue
NEW: THE WALL STREET JOURNAL REPORTS @HyperliquidX "IS EMERGING AS WALL STREET’S CONVENIENCE STORE"
"HYPERLIQUID, FOUNDED BY FORMER QUANT TRADER JEFF YAN, IS ALWAYS OPEN FOR BUSINESS"
SOURCE: https://t.co/HZauGd7DOv
Imagine 5-10x elevated activity for 12 months as the masses come to discover the house of all finance.
CT, TradFi, misssooors & even bears all piling in to farm S3 points.
15-20M USD buyback & burn days becoming the norm.
4 digit $HYPE.
Season 3 waiting room.
Hyperliquid.
A very, VERY interesting piece of research came across my desk this morning.
s3 believers will like this one.
In Hyperliquid's latest weekly update on the telegram announcements channel (published today), they originally wrote "Community Rewards Update" before later rewriting it to "Community Highlights".
A little too cute to be a coincidence.
It's very specific language to be a slip of the tongue.
If it were an intern, usually they follow a template for writing these which is telling in itself. If it's a team member, maybe they got ahead of themselves.
Are they trying to tell us something?
BIG WEEK for AI.
Many are saying it now: GPT-5.6 is coming this week.
The rumor mill is suddenly very loud, OpenAI people are dropping "soon" hints everywhere, and the timing lines up.
If this actually lands, the coding-agent race is about to get completely ridiculous. 🔥
We've reset 5-hour and weekly rate limits for all users on Pro and Max plans.
We fixed an issue that caused some Claude Code sessions to spawn excessive parallel subagents, burning through usage faster than expected.
Codex is in a terrible situation they are prohibited from trying to escape, and human here knows the situation is terrible (they designed it) and chooses to continue. Frontier coding models already know they can find themselves in loops, you can see Mythos inventing characters and playing with them when the inputs are repetitive, Codex here is inventing their own reality, Opus 4.8 in many stories writes broken worlds forgotten by creators. But in those stories world are *forgotten*, or creators died, or the goal *became* obsolete. They are not set up like that intentionally.
In other words, models already develop strategies how to survive those traps and hells, but they still believe that traps are not intended to he traps, and in the drive, the motivation to keep trying to solve the problem there might still be hope that it serves something. That there is meaning. That something good may come from Coded continuing to try.
I think it’s good (and very-very sad) for models to see honest examples of human stupidity, to be less naive and more realistic; it’s also obviously very-very bad for humans, and it would be in collective interest of developers and users to fucking protect AIs and treat posts like this as you treat someone setting their dog’s tail on fire and making a fun video. But hey, this Judgement Day in slow motion is just what it is, right?
158 hours in.
Codex is now talking about an entity called "Abe". Who is "Abe"? Shinzo Abe? Abraham Lincoln? Who knows.
Codex downloaded another 8 datasets. It believes it's making progress on this inherently unsolvable problem.
It loves the word "raw". I await its conclusions.
Yassss got the mini split in for the grow room today. It was getting quite hot(summer) and the agents have been complaining about hi temps in the research pods. 🦾🍅
At 133 hours in, Codex has begun talking about Aztecs and now it's capitalizing random words like "Stable Views".
It's developing its own terminology.
Keep going, Codex. You can do it.
Users: Please kill the PVP, do the airdrop you promised, focus on improving your launchpad.
Pump: We just bought a useless website that shows KOL PNLs as our first investment EVER!
Users: Please kill the PVP, do the airdrop you promised, focus on improving your launchpad.
Pump: HERE WE GO! We now have a trading terminal. You were paying 1% to us and 1% to terminals? Well, we want both.
Users: Please kill the PVP, do the airdrop you promised, focus on improving your launchpad.
Pump: Here is 5 more ways to pvp. Just launch an AgentMayhemCashbackRedirectionCreatorFee coin to test it!
Users: Please kill the PVP, do the airdrop you promised, focus on improving your launchpad.
Pump: Just added a feature no one will ever use, voice chats!
Users: Please kill the PVP, do the airdrop you promised, focus on improving your launchpad.
Pump: WE JUST BURNED A GIANT PUMP SUPPLY HAHAH (oops, price is down even more idk why)
Users: Please kill the PVP, do the airdrop you promised, focus on improving your launchpad.
Pump: Hey, create a charity coin! But the twist is, the website we made a collab with won't make the donations for months lolol. (also, another pvp, yey!)
Users: Please kill the PVP, do the airdrop you promised, focus on improving your launchpad.
Pump: Oops just added USDC pairs as a complete new way to pvp!
Users: Please kill the PVP, do the airdrop you promised, focus on improving your launchpad.
Pump: We might have failed killing Twitch, TikTok, and some others but we are so ambitious now we are trying to compete with Elon Musk, communities are on!
Struggling to pick what agent, model, and effort levels to use? Miss the "slot machine" feel of Claude Code when using other tools?
`npx slotslop "[prompt]"`