Akito

123 posts

Akito banner
Akito

Akito

@KonoAkito

Building @colosseum Frontier Hackathon Co-Founder @RootLens_sol Member @SuperteamJapan / Member @sol_hack3rs / Kobe Univ

Superteam_Japan Katılım Mayıs 2024
285 Takip Edilen131 Takipçiler
Sabitlenmiş Tweet
Akito
Akito@KonoAkito·
This explains the technology behind the anti-deepfake cameras of @RootLens_sol in an easy-to-understand way. I’d like to mention some people who expressed interest in RootLens. @k5cryptojp @by_olesia @mrinko @SuperteamJapan
RootLens | Frontier Hackathon@RootLens_sol

Introducing Rootlens, the anti-deepfake camera This is the foundational technology for combatting deepfakes across all sectors This won the Grand Prize at the @sol_hack3rs Global Hackathon, and our current entry is for the @colosseum Frontier Hackathon Product demo coming soon @C2PA_org

English
2
0
3
482
Dee
Dee@DaffyDudzz·
@KonoAkito @RootLens_sol @moodai0119 we sure are and I love that y’all built this, definitely see a lot of use cases for this . keep going, see y’all in the arena too though 😅
English
1
0
1
16
Akito
Akito@KonoAkito·
@jindu_ GM building 💪💪
English
0
0
0
17
Akito retweetledi
Hisa
Hisa@hisa_hack3rs·
Sol Hack3rsは常に日本に留まらず、世界に拡大し続けています。 Colosseum Hackathonにチャレンジしてみたい? 連絡してください。 僕らと一緒にハッカソンやメディアを運営してみたい? 連絡してください。 僕たちは世界中どこにいても情熱を共有する仲間をいつでも歓迎しています!
Sol Hack3rs@sol_hack3rs

the world' first student-led Solana hackathon co-hosted with @SuperteamJapan. 18 teams shipped 60+ founders, enterprises, investors & VCs $4k in prizes backed by @SolanaFndn, @minnano_ginko, @SlashWeb3 & more. they trained in Shibuya. now they're entering @colosseum.

日本語
1
4
12
1.2K
MonkeFoundry
MonkeFoundry@MonkeFoundry·
Applications for MonkeFoundry x Solana Demo Day close today. If you’re building on @Solana, this is an opportunity to share your work in a room of investors, operators, and ecosystem partners during Accelerate week in Miami. We’re bringing together a small group of teams we believe in.
MonkeFoundry tweet media
English
10
6
44
1.9K
HAKOBUNE栗島/ AIバーチャルカンパニー構築中
【ゆる募】VCアタックリスト生成AIが完成したのでβテスターとして活用したい起業家がいたらご連絡ください。コストの推移とか見たいので1日50件上限にしてます。アタックリスト生成とは別に投資家DBも実装してみてます
HAKOBUNE栗島/ AIバーチャルカンパニー構築中 tweet media
日本語
9
1
63
7K
Akito retweetledi
k5
k5@k5cryptojp·
really interested in this recent AI contents are hard to distinguish those are real or not this product may be able to provide proofs of pic/video are real. I have no idea what the standard will be but this kind of proofs will be necessary in very near future
RootLens | Frontier Hackathon@RootLens_sol

Introducing Rootlens, the anti-deepfake camera This is the foundational technology for combatting deepfakes across all sectors This won the Grand Prize at the @sol_hack3rs Global Hackathon, and our current entry is for the @colosseum Frontier Hackathon Product demo coming soon @C2PA_org

English
0
2
6
356
Akito
Akito@KonoAkito·
@256hax よろしくお願いします🙇
日本語
0
0
1
18
hax - ハックス
hax - ハックス@256hax·
RootLens、ディープフェイク対策のプロジェクトです。 KOLから戦場ジャーナリストまで、幅広くピッチのフィードバックをもらいながら、プロダクトを磨いているチームです。 勉強になる事実を教えてもらいました。 紛争地帯では情報戦になるため、フェイク画像や動画を使って混乱させるそうです。
RootLens | Frontier Hackathon@RootLens_sol

Introducing Rootlens, the anti-deepfake camera This is the foundational technology for combatting deepfakes across all sectors This won the Grand Prize at the @sol_hack3rs Global Hackathon, and our current entry is for the @colosseum Frontier Hackathon Product demo coming soon @C2PA_org

日本語
2
1
11
1.5K
Akito
Akito@KonoAkito·
@k5cryptojp @RootLens_sol @by_olesia @mrinko @SuperteamJapan 写真の内容自体はcNFTと紐づいてないので、他者から見えないようにできます 写真から抽出されるハッシュ値をcNFTと紐づけて記録してるので、写真自体を記録してなくても、本物証明に対応していた写真をすり替えるということはできないです
日本語
1
0
0
44
Akito
Akito@KonoAkito·
This explains the technology behind the anti-deepfake cameras of @RootLens_sol in an easy-to-understand way. I’d like to mention some people who expressed interest in RootLens. @k5cryptojp @by_olesia @mrinko @SuperteamJapan
RootLens | Frontier Hackathon@RootLens_sol

Introducing Rootlens, the anti-deepfake camera This is the foundational technology for combatting deepfakes across all sectors This won the Grand Prize at the @sol_hack3rs Global Hackathon, and our current entry is for the @colosseum Frontier Hackathon Product demo coming soon @C2PA_org

English
2
0
3
482
Akito
Akito@KonoAkito·
the difference comes down to trusting people versus trusting machines. While Numbers Protocol records data on the blockchain to prevent tampering, this protection only kicks in after the record is created. During the registration process, the verification of whether content is authentic is still handled by their internal servers. This means you have to trust the company to stay honest—if they wanted to, they could technically alter the data before it ever hits the blockchain. RootLens takes security a step further by moving that entire verification process into a Trusted Execution Environment (TEE)—an isolated hardware space that is completely cut off from external interference. Because the verification happens in this "black box," even we, the operators, cannot tamper with the content during the recording phase. Furthermore, because RootLens is open-source, anyone can verify for themselves that the processing was handled by the TEE rather than our own servers. This ensures a level of transparency and "math-based trust" that goes beyond just taking our word for it.
English
0
0
0
29
Akito
Akito@KonoAkito·
Spot on. We’re moving the battle from pixels to silicon. But here’s the thing: hardware attacks don’t scale as easily as software ones due to the sheer cost. By implementing a system where compromised hardware can have its trust revoked immediately, we ensure that any successful breach is both expensive and short-lived.
English
0
0
0
4
Saeed Anwar
Saeed Anwar@saen_dev·
@RootLens_sol C2PA binding at capture is genuinely interesting, the proof-of-real angle sidesteps the cat and mouse of deepfake detection. Main risk is that secure element supply chain attacks become the new attack surface once the economics justify it.
English
1
0
2
18
RootLens | Frontier Hackathon
📷 RootLens — Week 2 update We're building the anti-deepfake camera. Snap your day. Train AI. Get your pay. Just by shooting with your phone, your video becomes AI training data you can sell. The person who captured it gets paid. The problem Robotics companies are spending over $100M/year on real-world video data. They're strapping iPhones to gig workers' foreheads in Nigeria and India, paying $15/hour to film them doing chores. (MIT Tech Review, April 2026) But once you pay people per video, some will cheat by submitting AI-generated fakes to get paid. This is a problem across AI training data, already seen in text and images. If robots train on fake data, they fail in real homes. Our approach RootLens signs your video at the moment of capture, inside a secure chip. The signature binds sensor data like IMU and depth into the video through the C2PA standard, so the whole package is tamper-evident. Other sensor data can be added when it helps. We don't detect fakes. We prove what's real. This week we pivoted here. What we did ・Redefined our value proposition ・Deep-dived the Physical AI data market (competitors, regulation, tech stack) ・Rewrote our core assumptions ・Drafted the concept spec (system architecture, bounty mode, x402 settlement, KYC, EU AI Act) ・Built the pitch deck and pitch script ・Had a conversation with Shogo Omori (CAEN Inc.) Next week ・Start looking for pilot partners ・Iterate on the pitch through review rounds ・Publish technical breakdown videos on X Building in public. Feedback welcome.
English
1
0
9
554
Muse
Muse@muse_jp_sol·
ハッカソンのせいで寿命30年くらい縮んでると思う
日本語
3
0
16
822
Pedr◎
Pedr◎@Pedromiranda·
Solana
Español
99
5
342
10K