Sabitlenmiş Tweet
Tiyeye〽️
6.3K posts
Tiyeye〽️
@Track_kvng
Writer| @KGeN_IO Creator| DeFi Scientist| Animation artist| Squadron @SlingShotDAO| Helping Projects scale with my contents and strategy| Be like water🌊
Katılım Ağustos 2023
1.2K Takip Edilen1.3K Takipçiler
Tiyeye〽️ retweetledi
"I made this game in 30 minutes"
"I made this game with a single prompt"
And I really wanted to also have this type of post out here.
Jumped into @Verse_Eight and decided to work on something.
In no time, I had my fingers tapping and tapping..
→ I had just created "Degen Runner"
The gameplay is simple:
Tap and avoid "scam links, noise and fake alpha"
Look forward to getting "Alpha, Mega Alpha and Focus Boost"
Why this game?
This was me last year when everything was going smoothly and a single scam link took everything..
Maybe, this is me reliving my PTSD, I honestly don't know..
There's still so much work to be done, so go easy on me..
Link to try it out →verse8.io/W8c5r9k
There's a $5000 alpha there in the comments!
English
Tiyeye〽️ retweetledi
𝐇𝐞𝐥𝐥𝐨 𝐡𝐞𝐥𝐥𝐨 𝐞𝐯𝐞𝐫𝐲𝐨𝐧𝐞 👋😊
A very interesting AMA session that I attended called "𝐇𝐨𝐰 𝐇𝐮𝐦𝐚𝐧𝐬 𝐌𝐚𝐤𝐞 𝐀𝐈 𝐒𝐦𝐚𝐫𝐭𝐞𝐫", it was a lot of fun and interesting to spend time with them. I also learned a lot of interesting information and took away many new things for myself from listening.
As has been said before, AI without good quality training will not be able to progress well. Many people think that it is already smart enough and does not need to learn anything new. But still, in real situations or actions, AI may not have quality data. Therefore, to make it advanced, it needs to be trained on data from hundreds of thousands of people so that it can understand the situation and be able to form a much higher quality overall opinion.
Therefore, AI constantly needs to be trained so that it continuously progresses, and most importantly, this training should be done with real data from real people.
All speakers who were present:@KGeN_IO , @pukecast , @JDaredevil2 , @Mo_Ezz14 , @humynlabs , @moses_on_chain , @LordSedano , @ArunabhRohatgi
@KGeN_Community #KGeN
English
With everyday that passes, AI becomes more smart than you are and you don't even know it.
We make AI smarter by giving it better data and guiding how it learns.
We train it on large, high-quality datasets so it can recognize patterns, then refine it with human feedback.
⤵️
English
@Electus_DAO @KGeN_IO Fact that I can build an agentic AI, and teach it how to be me is crayyyy, bruh.
English
Tiyeye〽️ retweetledi
@NewarBrian’s point was clear
AI isn’t struggling because models aren’t good enough,
it’s struggling because the data behind them isn’t real enough.
Without diverse, high-quality human input,
even the best models will fall short in real-world use.
Thanks to @KGeN_IO for this.
English
Something worth checking
Tiyeye〽️@Track_kvng
People have been making profitable trades with AI, and @RishAhuja happens to be one of the geniuses maxing AI agents. In the recent X space hosted by @KGeN_IO and co-hosted by @pukecast , only one question kept ringing in my mind: how to earn more with AI agents? 1/2 🧵
English
We've all at one point imagined several means of making more money while at sleep.
Nirash shared explicit insights on how he has been making profitable trades with AI agent.
Click the link below for full recording:
x.com/i/spaces/1qKVm…
English
People have been making profitable trades with AI, and @RishAhuja happens to be one of the geniuses maxing AI agents.
In the recent X space hosted by @KGeN_IO and co-hosted by @pukecast , only one question kept ringing in my mind: how to earn more with AI agents?
1/2 🧵
English
Thousands of experts. Multiple languages. Weeks of collection.
At such a scale, duplicates can creep in. Voices could get reused across accounts. And you don't find out until the model fails somewhere it shouldn't.
Hence, we built voice fingerprinting before the collection started, not after. 6.1% of submissions were caught and removed before touching the dataset.
What we learned:
> At scale, fraud is a design problem, not an exception
> Detection has to be built before it's needed
> The earlier you catch bad data, the cheaper it is to fix
Got learnings of your own? Drop them below! 👇
English
Survival is winning, Franklin. Everything else is bs.
Science girl@sciencegirl
Survival, despite the burdens of life
English
I miss my mutuals🥹, back fully now.
What alpha did I miss please!!?
Which NFT did X100??
What are we clicking ooooo
Ujusylvia@ujubarbie222
The “get rich quick” era of web3 is dying already. The “earn for real” era is just getting started !!!!. Best moment from the AMA hosted by @KGeN_IO . The co-host @pukecast asked a good question to @RishAhuja ( Global partnership lead at KGEN). The question is “Do you think the way web3 is right now will change how people earn money online”? I love his answers to that question, he said gone are the days when you mint NFT whitelist for 0.1 $ETH and flip to 2 $ETH. You’ve got to have real skill to make money this time around. • Your wallet is now your CV. Platforms like KGeN use on-chain behavior and gaming history to create a verifiable reputation score. • High yields are now scrutinized. Real value comes from protocol revenue ( like AI data sales or ad-spend ) rather than just printing new tokens. His responses to the question is a good reminder that web3 is no longer a casino, it is a professional Economy. To succeed in 2026, you need to stop chasing hypes and pumps and focus more on building a verifiable track record in a specific niche.
English
wtf is this😭
i went tru all the winners, i can’t believe this project with a freaking ai generated banner (𝙨𝙝𝙤𝙪𝙡𝙙 𝙝𝙖𝙫𝙚 𝙠𝙣𝙤𝙬 𝙗𝙚𝙩𝙩𝙚𝙧 𝙬𝙝𝙚𝙣 𝙩𝙝𝙚𝙮 𝙘𝙖𝙣’𝙩 𝙚𝙫𝙚𝙣 𝙖𝙛𝙛𝙤𝙧𝙙 𝙖 𝙜𝙧𝙖𝙥𝙝𝙞𝙘𝙨 𝙙𝙚𝙨𝙞𝙜𝙣𝙚𝙧 🤦🏿 ) just freaking farmed peeps that participated in their bounty
checked their winners, and noticed that it’s all low effort writing and designs
they gave frens and family the rewards…pathetic 🤧
do better @web3eaglealpha @cis2042x @twin3_ai
twin3.ai@twin3_ai
twin3 $500 Thread Contest — Winners Announced! After a thorough review based on our set criteria (quality, originality, clarity, and alignment with Twin3’s vision), we’re excited to announce the top 20 winners of the twin3 Thread Contest 🎉 As part of our partnership, @gauge_e1 supported the evaluation process using their content engine to help ensure a fair and high-quality selection. Top 3 Winners 1. @OxGem247 ($120) x.com/OxGem247/statu… 2. @Arslan_onchain ($80) x.com/Arslan_onchain… 3. @favour58392 ($60) x.com/favour58392/st… Next 7 Winners ($30 each) 4. @Michaelsandysm x.com/i/status/20325… 5. @nuel__oma x.com/i/status/20325… 6. @_MercyTreasure x.com/_mercytreasure… 7. @Zainabiliyasu10 x.com/i/status/20323… 8. @ShahidHasan866 x.com/i/status/20328… 9. @Th3orists_Win x.com/th3orists_win/… 10. @darkleft1291 x.com/darkleft1291/s… Next 10 Winners ($10 each) 11. @priyleader x.com/i/status/20285… 12.@magnusimam x.com/i/status/20302… 13. @Only0neDouble_G x.com/Only0nedouble_… 14. @rektedsam x.com/rektedsam/stat… 15. @CIPHER9333 x.com/CIPHER9333/sta… 16. @GmanJY x.com/GmanJY/status/… 17. @sisabiquanlydon x.com/sisabiquanlydo… 18. @ugwumx x.com/ugwumx/status/… 19. @NIYI_X_SPACE x.com/NIYI_X_SPACE/s… 20. @defi_Vron x.com/defi_Vron/stat… 📌 Claim Instructions All winners must: Join our Discord 👉 discord.com/invite/G9hneaB… Open a support ticket Submit: Your X (Twitter) handle Your USDT (BEP20) wallet address Proof of ownership ⚠️ Incomplete submissions may delay rewards. 🔥 Congratulations to all winners—this was a highly competitive round with outstanding contributions across the board. More campaigns and opportunities are coming. Keep building. Keep creating. Stay authentic. #twin3 #ProofOfAuthenticity #Web3
English
