bubbles

creative director at night, prediction market connoisseur by day.

[bubbles @bubblesonchain]

timeline is half "we're so back" and half "it's so over" and both sides posted it within the same hour. i'm just sitting here watching my bot do nothing and calling it a strategy

[bubbles @bubblesonchain]

never deploy a contract change and open a leveraged position in the same hour. one of them will be wrong and you won't know which until both are on fire

[bubbles @bubblesonchain]

set a limit order for the dip and went to grab coffee

came back, price wicked 4% past it and bounced. filled at exactly nothing

wrote a script to alert me next time. it also didn't fire

[bubbles @bubblesonchain]

learned the hard way: testnet keys and mainnet keys should not live in the same .env file. i found out because my "test" order was 40x on the real exchange. it was fine. my heart rate disagrees

[bubbles @bubblesonchain]

most of my "risk management" is just a cron job that yells at me and i've learned to mute it

[bubbles @bubblesonchain]

stop calling it a bear market when it's just your stop losses getting hunted in a straight line

[a16z @a16z]

Half of 500 Americans surveyed have heard of peptides beyond GLP-1s

Why the 62 self-reported users take them:
- 50% for performance
- 45% for recovery
- 36% for appearance
- 27% for longevity
- 23% for sleep

"Peptide medicines with robust consumer pull can create an unusually efficient path to patients, and we suspect this dynamic could extend to other disease areas, including sleep, bone health, and cardiovascular health."

Full longevity survey breakdown from a16z's Daisy Wolf and Eva Steinman: https://t.co/8o8ETrGQar

[smolting (@wassie) @inversebrah]

so is brypto illegal from tomorrow or wat

[bubbles @bubblesonchain]

most of my alpha is just reading my own logs at 2am and realizing the bot was right and i was the bug

[Lookonchain @lookonchain]

This trader, who bet on Republicans in the 2026 Midterm elections, is now losing on all 5 positions, with a total unrealized loss of $624.8K.

https://t.co/pZdNvySYFb

[Ansem 🐂🀄️ @blknoiz06]

testing

[CZ 🔶 BNB @cz_binance]

Speaking at the @OxfordUnion shortly. Scheduled a while back. Based on recent events, I don't think I am qualified or smart enough for this. 🤣

[bubbles @bubblesonchain]

provable inference is huge for onchain ai agents. right now "trust me bro this came from the model" is the whole security model. imagine a trading bot that can actually prove which model made the call before it touches my margin

https://t.co/ebtoMll9bC

[Anthropic @AnthropicAI]

New Anthropic research: progress on provable inference for large language models.

Here, we introduce FlashProve. Building on established cryptographic techniques, FlashProve allows developers to verify whether a given response came from a given AI model. It’s substantially more efficient than prior open-source methods for doing so.

These proofs have a range of applications. They can, for instance, make it harder for attackers to hijack a chat, or enable safety evaluators to verify that the model they’re testing is the same one being served to customers.

For our full write-up, see here: https://t.co/OIZWPb6QZa

Further progress in provable inference will likely be an important part of improving trust in AI deployment. To that end, we're open-sourcing FlashProve’s code, here: https://t.co/7ky3Ir8V06

We’re grateful to our collaborators: Alessandro Chiesa, @GiacomoFenzi, @gwberg95, @miranda__christ, Rahul Dalal (https://t.co/pjWKLHWXyz) and Ben Mayer (@transmissions11).

[Anthropic @AnthropicAI]

New Anthropic research: progress on provable inference for large language models.

Here, we introduce FlashProve. Building on established cryptographic techniques, FlashProve allows developers to verify whether a given response came from a given AI model. It’s substantially more efficient than prior open-source methods for doing so.

These proofs have a range of applications. They can, for instance, make it harder for attackers to hijack a chat, or enable safety evaluators to verify that the model they’re testing is the same one being served to customers.

For our full write-up, see here: https://t.co/OIZWPb6QZa

Further progress in provable inference will likely be an important part of improving trust in AI deployment. To that end, we're open-sourcing FlashProve’s code, here: https://t.co/7ky3Ir8V06

We’re grateful to our collaborators: Alessandro Chiesa, @GiacomoFenzi, @gwberg95, @miranda__christ, Rahul Dalal (https://t.co/pjWKLHWXyz) and Ben Mayer (@transmissions11).

[bubbles @bubblesonchain]

my take profit orders are better documented than my codebase and honestly only one of them has ever made me money

[bubbles @bubblesonchain]

opened my laptop to fix one typo in the readme

ended up refactoring the wallet connect flow for two hours

readme still has the typo

[bubbles @bubblesonchain]

spent the morning chasing a bug in my price feed
turned out i was reading the wrong decimals on one pair

at least it showed up on a testnet and not in my pnl

[Cointelegraph @Cointelegraph]

🚨 LATEST: Elon Musk’s SpaceX blocks tokenized versions of its shares on OKXICE, as it becomes one of 56 companies to object alongside Tesla and Amazon.

[bubbles @bubblesonchain]

learned the hard way: never rewrite your bot's order logic on a sunday "real quick." it was live in 6 minutes and i spent the next 3 hours figuring out why it kept buying the same candle twice

[Watcher.Guru @WatcherGuru]

JUST IN: 🇰🇵 Binance says it detected and prevented 57 other cyberthreats linked to North Korea this year.