kab retweeted
Brands spend weeks perfecting a single ad, only to need ten more the moment it goes live.
@nynaoai takes your brief, learns your brand, and ships finished video ads every day,no editors in the loop.
You can’t just sign up. You have to apply.
Which, honestly, makes me want it more.
🤝 Paid partnership
It's Q4.
You hired 3 editors, then 4, then 10. And when you need volume, they still can't keep up.
That's why we built Nynao.
Up to 400 a day with no editors.
Apply for access : nynao.com
here are cyber attacks and major outages that has happened in november 2026 alone with major aids from ai
— november 2: european supermarkets (carrefour, ikea, ica, etc.) – refrigeration systems hacked.
— november 12–13: microsoft outlook – worldwide service outage.
— november 16: jpmorgan chase, usa – banking systems outage.
— november 16: maersk, denmark – reported data breach.
— november 16: tsmc, taiwan – employee data exposure.
— november 16: osaka city government, japan – email data exposure.
— november 16: casinos austria – third-party data breach.
— november 16: optimizely – supply-chain cyberattack.
— november 17: us national security agency (nsa) – classified communications network outage.
— november 17: equinor, norway – ongoing cyberattack.
and we're only just getting started
kab retweeted
Found a free AI tool that lets you generate unlimited videos without paying a dime.
I broke down how to use it.
Save this, you’ll thank me later.
kab retweeted
This guy shows you in 22 minutes how to learn to create videos automatically with Claude Code
Save this.
i'll be honest, whenever someone talks about private transactions, my first question is always the same. how long does it take for a transaction to go through?
that has been one of the biggest problems with FHE. even a simple 64-bit transfer gets broken down into thousands of tiny logic gates, and each one comes with its own bootstrapping process. that makes things incredibly slow.
@primus_labs is trying to change that with autohog. instead of processing all those small gates separately, it combines them into larger ones that can share a single bootstrapping pass.
here are some of the numbers they're reporting:
→ around 90ms per confidential transfer
→ 210+ transfers per second
→ running on 8 consumer rtx 5090 gpus
→ identical outputs across every node, byte for byte
the part i find interesting is the use of consumer gpus. it means this kind of computing may not require expensive, specialized hardware or a handful of heavily funded operators.
i've also had the chance to try primus myself. i've verified my pnl without exposing my wallet, and tonight i put a small amount into their confidential vault. the yield stays the same, but my position isn't publicly visible.
now think about what this could mean for lending, funds and other onchain activities where people don't necessarily want every move exposed to the public.
of course, these are still benchmark results. the real test is seeing how well they hold up on mainnet with actual users and traffic.
but if the performance carries over, private onchain finance could finally become something people use rather than just another interesting idea.
explore here: s.kaito.ai/m1ePRuT
🤝 Paid partnership
everything onchain is public. that's great for trust, but it's a real problem for institutions.
no fund wants competitors watching its positions, cash flow and strategy in real time. so big money stays on the sidelines. it needs proof that can be verified and data that stays private, both at once.
that's the gap @primus_labs is going after, with two tools:
zkTLS proves that data from a website or API is real, like an exchange balance, without showing the exact numbers.
FHE lets smart contracts work with encrypted data, so sensitive info never gets exposed.
and it's already moving. Confidential Vault is live on BNB Chain, the extension has passed 1M users, and Nillion and Tickr use zkTLS to verify Binance data privately. Primus has also raised $6.5M, backed by Dispersion Capital, Symbolic Capital and VanEck.
the idea is confidentiality, not anonymity. hide the amounts, keep the truth verifiable.
as stablecoins and tokenized assets keep moving onchain, that's what institutions will ask for.
you can follow this closely here: s.kaito.ai/m1ePRuT
gm ct...
happy wednesday
i'll be honest, whenever someone talks about private transactions, my first question is always the same. how long does it take for a transaction to go through?
that has been one of the biggest problems with FHE. even a simple 64-bit transfer gets broken down into thousands of tiny logic gates, and each one comes with its own bootstrapping process. that makes things incredibly slow.
@primus_labs is trying to change that with autohog. instead of processing all those small gates separately, it combines them into larger ones that can share a single bootstrapping pass.
here are some of the numbers they're reporting:
→ around 90ms per confidential transfer
→ 210+ transfers per second
→ running on 8 consumer rtx 5090 gpus
→ identical outputs across every node, byte for byte
the part i find interesting is the use of consumer gpus. it means this kind of computing may not require expensive, specialized hardware or a handful of heavily funded operators.
i've also had the chance to try primus myself. i've verified my pnl without exposing my wallet, and tonight i put a small amount into their confidential vault. the yield stays the same, but my position isn't publicly visible.
now think about what this could mean for lending, funds and other onchain activities where people don't necessarily want every move exposed to the public.
of course, these are still benchmark results. the real test is seeing how well they hold up on mainnet with actual users and traffic.
but if the performance carries over, private onchain finance could finally become something people use rather than just another interesting idea.
explore here: s.kaito.ai/m1ePRuT
kab retweeted
AI video creation just leveled up
Someone dropped two new AI video tools and you can use both without spending a single dime
Save this so you don't lose it
here are cyber attacks and major outages that has happened in november 2026 alone with major aids from ai
— november 2: european supermarkets (carrefour, ikea, ica, etc.) – refrigeration systems hacked.
— november 12–13: microsoft outlook – worldwide service outage.
— november 16: jpmorgan chase, usa – banking systems outage.
— november 16: maersk, denmark – reported data breach.
— november 16: tsmc, taiwan – employee data exposure.
— november 16: osaka city government, japan – email data exposure.
— november 16: casinos austria – third-party data breach.
— november 16: optimizely – supply-chain cyberattack.
— november 17: us national security agency (nsa) – classified communications network outage.
— november 17: equinor, norway – ongoing cyberattack.
and we're only just getting started
gm ct...
happy wednesday
i'll be honest, whenever someone talks about private transactions, my first question is always the same. how long does it take for a transaction to go through?
that has been one of the biggest problems with FHE. even a simple 64-bit transfer gets broken down into thousands of tiny logic gates, and each one comes with its own bootstrapping process. that makes things incredibly slow.
@primus_labs is trying to change that with autohog. instead of processing all those small gates separately, it combines them into larger ones that can share a single bootstrapping pass.
here are some of the numbers they're reporting:
→ around 90ms per confidential transfer
→ 210+ transfers per second
→ running on 8 consumer rtx 5090 gpus
→ identical outputs across every node, byte for byte
the part i find interesting is the use of consumer gpus. it means this kind of computing may not require expensive, specialized hardware or a handful of heavily funded operators.
i've also had the chance to try primus myself. i've verified my pnl without exposing my wallet, and tonight i put a small amount into their confidential vault. the yield stays the same, but my position isn't publicly visible.
now think about what this could mean for lending, funds and other onchain activities where people don't necessarily want every move exposed to the public.
of course, these are still benchmark results. the real test is seeing how well they hold up on mainnet with actual users and traffic.
but if the performance carries over, private onchain finance could finally become something people use rather than just another interesting idea.
explore here: s.kaito.ai/m1ePRuT
kab retweeted
𝗰𝗮𝗻 𝗳𝘂𝗹𝗵𝗮𝗺 𝘁𝗮𝗸𝗲 𝗮𝗱𝘃𝗮𝗻𝘁𝗮𝗴𝗲 𝗼𝗳 𝗺𝗮𝗻 𝗰𝗶𝘁𝘆'𝘀 𝘃𝘂𝗹𝗻𝗲𝗿𝗮𝗯𝗶𝗹𝗶𝘁𝘆?
city plays fulham on saturday, and i'm not as confident about this one as the odds suggest.
here's why:
➔ city lost 3-1 to nottingham forest in their last league game.
➔ they've conceded in 4 of their last 5 matches.
➔ fulham have picked up points in 4 of their last 5 league games.
➔ city have won their last 10 meetings with fulham.
➔ fulham have scored in 4 of their last 5 matches.
@bagel_win has city at 70c, meaning the market gives them a 70% chance of winning.
i understand why city are favourites, but 70c feels a little too expensive for me.
would you trust city at 70c, or is fulham worth a look?
🤝 Paid partnership
kab retweeted
THIS IS INSANE🤯
I just found an 8 minute guide that shows you how to code and build apps with Claude.
it breaks down the process into simple steps
definitely worth saving.
To learn AI automation for free check here drive.google.com/drive/mobil……
To learn Digital Marketing for free check here drive.google.com/drive/mobil……
To learn UI/UX design for free check here
drive.google.com/drive/mobil……
For AI YouTube automation check here drive.google.com/drive/mobil……
Repost, bookmark and get to work🔖
kab retweeted
ai is really advancing daily.
new and more powerful models keep dropping, major platforms are reporting outages and hackers are using AI to launch automated cyberattacks.
i just hope the disadvantages of AI don’t end up outweighing the advantages in the near future 😹
spreads can make or break a prediction market trade
you can be right about an outcome and still lose because of your entry price.
EventLiquid uses an AMM, a system that automatically provides liquidity and sets prices, with the goal of making markets deeper as more positions open.
expecting more cooks from the team, so I’d stay positioned
Wide spreads hold prediction markets back.
Meet EventLiquid: powered by LS-LMSR, an AMM that deepens liquidity as open positions grow.
More depth. Continuous prices. Always liquid.
Join the waitlist → eventliquid.com
kab retweeted
I used to waste way too much time checking random charts on my phone when I was away from my desk.
Moving my daily tracking to the @GeckoTerminal watchlist made things way easier.
Love how it puts real-time liquidity and volume across 2,000+ DEXs into one clean mobile view.
Right now, my list keeps $HYPE, $LAB, $PUMP, and $ANSEM side by side giving me an instant pulse on large-cap volume, BNB Chain liquidity, and Solana flow without opening a browser.
Check out my Watchlist in the cs
i'll be honest, whenever someone talks about private transactions, my first question is always the same. how long does it take for a transaction to go through?
that has been one of the biggest problems with FHE. even a simple 64-bit transfer gets broken down into thousands of tiny logic gates, and each one comes with its own bootstrapping process. that makes things incredibly slow.
@primus_labs is trying to change that with autohog. instead of processing all those small gates separately, it combines them into larger ones that can share a single bootstrapping pass.
here are some of the numbers they're reporting:
→ around 90ms per confidential transfer
→ 210+ transfers per second
→ running on 8 consumer rtx 5090 gpus
→ identical outputs across every node, byte for byte
the part i find interesting is the use of consumer gpus. it means this kind of computing may not require expensive, specialized hardware or a handful of heavily funded operators.
i've also had the chance to try primus myself. i've verified my pnl without exposing my wallet, and tonight i put a small amount into their confidential vault. the yield stays the same, but my position isn't publicly visible.
now think about what this could mean for lending, funds and other onchain activities where people don't necessarily want every move exposed to the public.
of course, these are still benchmark results. the real test is seeing how well they hold up on mainnet with actual users and traffic.
but if the performance carries over, private onchain finance could finally become something people use rather than just another interesting idea.
explore here: s.kaito.ai/m1ePRuT
🤝 Paid partnership
everything onchain is public. that's great for trust, but it's a real problem for institutions.
no fund wants competitors watching its positions, cash flow and strategy in real time. so big money stays on the sidelines. it needs proof that can be verified and data that stays private, both at once.
that's the gap @primus_labs is going after, with two tools:
zkTLS proves that data from a website or API is real, like an exchange balance, without showing the exact numbers.
FHE lets smart contracts work with encrypted data, so sensitive info never gets exposed.
and it's already moving. Confidential Vault is live on BNB Chain, the extension has passed 1M users, and Nillion and Tickr use zkTLS to verify Binance data privately. Primus has also raised $6.5M, backed by Dispersion Capital, Symbolic Capital and VanEck.
the idea is confidentiality, not anonymity. hide the amounts, keep the truth verifiable.
as stablecoins and tokenized assets keep moving onchain, that's what institutions will ask for.
you can follow this closely here: s.kaito.ai/m1ePRuT




























