Bitsec | Bittensor Subnet 60 τ

233 posts

Bitsec | Bittensor Subnet 60 τ banner
Bitsec | Bittensor Subnet 60 τ

Bitsec | Bittensor Subnet 60 τ

@bitsecai

AI Security Agents. Find and fix *all* exploits in your code

Katılım Mayıs 2024
42 Takip Edilen2.2K Takipçiler
Sabitlenmiş Tweet
Bitsec | Bittensor Subnet 60 τ
Bitsec | Bittensor Subnet 60 τ@bitsecai·
Can spending $1k to $5k / mo with @bitsecai save companies from $100k or more in catastrophic exploits? We are entering Phase 1. Here's the short summary of where we are and a taste of where things are going. We have 3 companies in the pipeline with a growing backlog of companies we still need to meet. Each new potential client we want to knock them out with a simple message: "Your codebase has a couple exploits ready to fix, have it for free whether you want to join or not. Subscribe and we'll spend more resources to do a thorough end to end exploration with our AI security army." We are exploring distribution channels, where closing one deal allows us to reach many ideal clients. EOY goal is 50 company subscriptions, 1 enterprise subscription. What this means is building a book of revenue between $50k to $250k MRR, building good reputation with paying clients, which can offset the current upfront inference cost per client to show them sample exploits. We need to add key personnel to the Bitsec team as we grow to service more clients. Phase 2 post coming soon. $Bittensor $TAO
Bitsec | Bittensor Subnet 60 τ tweet media
Bitsec | Bittensor Subnet 60 τ@bitsecai

x.com/i/article/2069…

English
4
9
72
8.4K
Bitsec | Bittensor Subnet 60 τ
Bitsec | Bittensor Subnet 60 τ@bitsecai·
"...you know, there’s one [subnet] for example that we have been close to since the very beginning, which is Bitsec, and man they did a great job at analysing the security of our code, and you know, kudos because they did find a very important vulnerability that we actually had to delay the release of it because of it. So we had to address that and we continue to monitor the security of all the code and everything with them." -Jose Caldera, Yanez founder/CEO We’re proud to announce our partnership with Jose and the talented Yanez team, one of the deals we secured during our recently completed sales Test Phase. And even though we just started Phase I, there are already several in the pipeline. Stay tuned. @bitsecai @yanez__ai $TAO @josercaldera @yubrew
Bitsec | Bittensor Subnet 60 τ tweet media
Bitsec | Bittensor Subnet 60 τ@bitsecai

x.com/i/article/2067…

English
1
3
20
576
Bitsec | Bittensor Subnet 60 τ
Bitsec | Bittensor Subnet 60 τ@bitsecai·
"It's not like the code base is going to stop. You're going to keep developing new features…" - John Yu, Bitsec Founder/CEO AI-generated code carries 2.74x more vulnerabilities than human-written code (Veracode). And code ships faster than any team can manually review A single audit before launch catches what existed at launch. It says nothing about what gets introduced six features later... and by then, it's already live. Security that stops when the audit ends isn't security. It's a snapshot. @bitsecai $TAO
Bitsec | Bittensor Subnet 60 τ@bitsecai

Bitsec beats Fable 5 Have evidence, publishing soon. Client oks release and verify to trusted 3rd party.

English
4
8
47
3.4K
Bitsec | Bittensor Subnet 60 τ retweetledi
TAO.com
TAO.com@taodotcom·
NEW: Bitsec says SN60 runs top subnet agents against codebases to clear known vulnerability classes, then focuses tokens on edge cases to surface novel exploits. x.com/bitsecai/statu…
Bitsec | Bittensor Subnet 60 τ@bitsecai

ELI5 Bitsec's security approach We start by running the top agents from our subnet competition on codebases, efficiently clearing out known landmines. Then, we spend the bulk of time and tokens combing through edge cases to detect new and novel exploits. In our testing, this approach generalizes well beyond blockchain projects. We find exploits others miss.

English
1
1
16
1.1K
Bitsec | Bittensor Subnet 60 τ
Bitsec | Bittensor Subnet 60 τ@bitsecai·
ELI5 Bitsec's security approach We start by running the top agents from our subnet competition on codebases, efficiently clearing out known landmines. Then, we spend the bulk of time and tokens combing through edge cases to detect new and novel exploits. In our testing, this approach generalizes well beyond blockchain projects. We find exploits others miss.
Bitsec | Bittensor Subnet 60 τ tweet media
Bitsec | Bittensor Subnet 60 τ@bitsecai

x.com/i/article/2072…

English
4
10
43
2.6K
Bitsec | Bittensor Subnet 60 τ retweetledi
YVR τrader
YVR τrader@YVR_Trader·
1990s IT security: “We installed antivirus and locked the floppy disks.” 2026 IT security: autonomous agents battling zero-day exploits across decentralized networks like @bitsecai #SN60 in real time. Same goal, wildly different battlefield.
YVR τrader tweet media
English
4
3
20
700
Bitsec | Bittensor Subnet 60 τ
Bitsec | Bittensor Subnet 60 τ@bitsecai·
Agent submissions is not free, it costs miners $100-$250 in inference per agent. With 138 agents submitted so far this round that is over $20k us dollars in inference. The current pool of miners is fairly competitive, with at least 3+ distinct groups. Revealing agents publicly every round pushes advancements every 2-3 weeks. Winner take all is cleaner from an IM design. The subnet designer does not need to worry as much about sybil attacks, judging criteria, screening criteria, plagiarism, similarity, gaming as strongly with one clear winner. Rewarding additional miners increases competitiveness, but it also increases surface area of attacking incentives.
peter fleming@peter_m_fleming

@bitsecai The success of the subnet depends on the miners i assume, im totally clueless i just invested cause i like the owner, but if you want fresh new talent would it not be better to pay down to top 5 or something , if same people keep winning what incentives do the rest have

English
1
2
36
3.7K
Bitsec | Bittensor Subnet 60 τ
Bitsec | Bittensor Subnet 60 τ@bitsecai·
Thanks to @markjeffrey for explaining Bitsec v Mythos so clearly on Novelty Search 9:30 timestamp We test new models to find new undiscovered vulnerabilities. It's a fair test on unseen codebases. If Fable 5 finds new things, it's added to our stack. Bitsec found more in this case. Everything Fable 5 uncovered, and additionally 5 criticals, 10 highs. youtu.be/OnrLSGH1HDw?si…
YouTube video
YouTube
Bitsec | Bittensor Subnet 60 τ tweet media
English
3
11
66
4.3K
Bitsec | Bittensor Subnet 60 τ
Bitsec | Bittensor Subnet 60 τ@bitsecai·
@peter_m_fleming It is winner take all, only the miner that made the top agent gets emissions. Compared to the rewards, it is not expensive to enter the competition. If your agent is much better you get $20k or more per round. We're experimenting to reward top two miners next round.
English
1
0
2
97
peter fleming
peter fleming@peter_m_fleming·
@bitsecai My only holding is SN60 , just curious about how it works, it shows two miners on taostats ,is that just because they are the only two getting emissions, does each miner have multiple agents that all share information, does winner take all not disincentivise new miners coming ?
English
1
0
1
78
Bitsec | Bittensor Subnet 60 τ
Bitsec | Bittensor Subnet 60 τ@bitsecai·
Must watch 👀 First time on HashRate and @markjeffrey is very welcoming and an excellent host
Mark Jeffrey@markjeffrey

Hash Rate - Ep. 177: Bitsec SN60 Beats Mythos(!) 🧙 Guest: @yubrew of @bitsecai A Bittensor $TAO Subnet Just Beat Anthropic's MYTHOS at Cybersecurity 00:00 Introduction to Bitsec 03:01 Jailbreaking Fable: Mythos Unleashed 12:02 The Competitive Landscape 17:49 The Future of Bitsec 30:03 John Yu's Journey into Cybersecurity 40:46 Subnet Token 44:53 Sales Strategy 49:40 Final Thoughts and Future Directions

English
1
10
68
2K
deploy
deploy@decoding_things·
@peter_m_fleming @bitsecai they’ve got a chance to become a lot of people’s favourite subnet if they execute well. can’t wait for further updates and/or dashboard
English
2
0
2
49
Bitsec | Bittensor Subnet 60 τ
Bitsec | Bittensor Subnet 60 τ@bitsecai·
Can spending $1k to $5k / mo with @bitsecai save companies from $100k or more in catastrophic exploits? We are entering Phase 1. Here's the short summary of where we are and a taste of where things are going. We have 3 companies in the pipeline with a growing backlog of companies we still need to meet. Each new potential client we want to knock them out with a simple message: "Your codebase has a couple exploits ready to fix, have it for free whether you want to join or not. Subscribe and we'll spend more resources to do a thorough end to end exploration with our AI security army." We are exploring distribution channels, where closing one deal allows us to reach many ideal clients. EOY goal is 50 company subscriptions, 1 enterprise subscription. What this means is building a book of revenue between $50k to $250k MRR, building good reputation with paying clients, which can offset the current upfront inference cost per client to show them sample exploits. We need to add key personnel to the Bitsec team as we grow to service more clients. Phase 2 post coming soon. $Bittensor $TAO
Bitsec | Bittensor Subnet 60 τ tweet media
Bitsec | Bittensor Subnet 60 τ@bitsecai

x.com/i/article/2069…

English
4
9
72
8.4K
deploy
deploy@decoding_things·
@bitsecai could you make a dashboard of live figures? you don’t have to put company names you’re working with just the amount of companies that have signed up and revenue per month even if it’s low now. would be awesome to track
English
1
0
3
91

Keşfet

@yanez__ai @josercaldera @yubrew @TAOTemplar @markjeffrey @peter_m_fleming @decoding_things @elonmusk