Trojan horse Wagyu / @PerpetualCow exit scammed on XMR / HL users?
Remember @TechnoRevenant promoted this guy.
There are many fake ".hl's" trying to cause harm to the community.
Wagyu was long removed from hlxyz FE's, yet they tried to cause harm to HL in any way they could. Now rugged.
There are some "HL maxis" that in reality are trojan horses trying to cause harm to the ecosystem.
@TechnoRevenant is one of them.
- XPL cascade
- kHYPE depeg attempt (raped himself)
- Sold 10M+ HYPE with CZ at a time to liquidate holders & gain more supply or destroy the community.
- lured the community into $aster
Now he tried to liquidate SM and cascade HYPE on the weekend CZ & Binance scammed the industry.
Techno is either CZ himself or CZ's dog on a leash.
Don't trust him or anyone he's endorsed (wagyu - also Trojan horse).
Hyperliquid.
not confusing why there is so many competitors already
@SpaceXSI hiding the fact of giving u shittier models than there already behind grok 4.7 all while usage is still absolute ass lololol
even crazier when u realize this is all after they false marketed the switch from legacy ultra cursor sub to new sub would be worth it for usage limits for @bot
32
quantum computing is todays version of string theory
Today Physical Superintelligence PBC is announcing a proof of the Aaronson-Ambainis conjecture, one of the central open problems in quantum complexity, with major implications for quantum computing speedup.
psi.inc/research/aaronson-am…
Thread 👇
1
41
QuestDB for recent data
export aged partitions nightly to Parquet
load those into ClickHouse for long-term storage and backtesting 🤝
I think AI has cracked coding but it has definitely not cracked Data Engineering.
There is nothing I can not build these days, but all the projects I am building are consuming, digesting and storing large amounts of data. And it's hard to manage that specially if you're self hosting like me.
Some of my DBs are growing 1Gb/day and Claude/Astra have still not yielded satisfactory results despite endless rounds of optimizations. It affects speed but also overall not sustainable in terms of storage
Im mostly rawdogging postgreSQL, so at some point will have to combine it with other solutions, the clankers keep suggesting clickhouse so it's something I'm looking at. But I'd like to keep it all self hosted.
If anyone has any tips im all hears.
3
47
xitter.cf/ChairmanSelig/status/2… Hyperliquid
Replying to @smoltide
Also I think perps legalization in US and becoming main stream is coming and @FoftyPawlow take will age terribly
All love tho big fan of him
1
1
68
this kinda stuff is useless at fixing context drift still....
an ironic take away is to realize bots just like humans do well with a couple rules and a single task they can repeat over and over..... but the more rules and tasks you add the worse the outcomes get
how one manages that is how you fix "context drift"....
real fix is simpler systems that keep the rules and responsibilities narrow, push the durable stuff into clear data structures and code wherever you can
i HATE when grok bot does this.
my bots slowly drift from the rules we set.
if you didn't know: anything you decide in chat fades over time. whatever's in the bot's profile is what it keeps in mind every turn.
my new bot had a whole morning of tweaks living only in chat. its profile still had the day-one brief.
what i'm doing about it:
1. every rule change gets saved to the bot's profile the moment we decide it.
2. a weekly routine in my chief of staff @bot compares what each bot actually did to its profile and flags drift with receipts. nothing changes without my ok
if you run this on your own bots, you might be surprised what you find tbh.
prompt is in the image.
3
78












