ethosevo

Deep tech & security R&D lab
Breaking systems to understand how to build them right.

[ethosevo @ethosevo]

🚀 Launching @ethosevo — a deep tech company I built from scratch.

🧠 Blockchain R&D (L2, ZKP, Web3)
🔐 Cybersecurity & Custom WAF
🤖 AI Infra & LLM Training
🛠️ OS + Language Design
☁️ Cloud-native Architecture

Pushing boundaries in AI & Web3 — one block at a time.

[ethosevo @ethosevo]

@DanielMiessler Strong take.

AI doesn’t replace engineers — it replaces passive execution.

People who only followed tickets get automated.
People who explore systems, build side projects, and understand failure modes gain leverage.
Tools change. Curiosity compounds.

[ethosevo @ethosevo]

@schickling This is exactly how real code review happens in production.

Diffs show what changed, but not what broke, what slowed down, or what regressed.

Reviews need runtime context: logs, traces, benchmarks, UI deltas, and history — not just lines of code.

[ethosevo @ethosevo]

Yesterday I shared a real production incident:
a Monero miner running silently on a hardened VM.
No SSH abuse.
No cron.
No persistence.

Entry point wasn’t infra — it was the app layer.
Lesson: most compromises bypass “traditional hardening”.
#AppSec #RCE #DevSecOps

[ethosevo @ethosevo]

Prod VM anomaly: fans nonstop.
top showed >300% CPU, high nice → xmrig miner.
No SSH abuse. No cron. No systemd/PM2 persistence.
PPID=1 → app-layer RCE (likely recent Next.js).

Lesson: SSH hardening ≠ app security. Fan noise is still an IDS.
#Security #RCE #NextJS #DevSecOps

[ethosevo @ethosevo]

@davj It’s not about farming.
It’s about wanting direct feedback loops again
effort → outcome, cause → effect.
Modern tech work abstracts that away, and the brain pushes back.

[ethosevo @ethosevo]

@imbktan This is the part most people miss.

AI didn’t replace understanding
it compressed the cost of getting to understanding.
The hard work is still specs, edge cases, and owning the system.

[ethosevo @ethosevo]

@SuhailKakar This is a usability and incentives problem, not a hypocrisy one.

Most crypto apps still feel like infrastructure demos, not products.
People talk because talking is cheaper than bearing UX, risk, and friction.

[ethosevo @ethosevo]

@luhelminger ZK isn’t dying
ZK-as-a-standalone identity is.
The survivors will be teams that embed ZK into real products, real workflows, and real distribution.
Tokens buy runway. Only integration buys relevance.

[ethosevo @ethosevo]

@ghoshal This isn’t a bug it’s an incentive mismatch.

When capital has no hard performance constraints,
the rational strategy shifts from building value
to maximizing optionality and downside protection.

[ethosevo @ethosevo]

@sjdedic Capital markets don’t reward truth.
They reward convincing narratives under uncertainty.
Product quality compounds slowly.
Storytelling arbitrages the gap before reality catches up.

[ethosevo @ethosevo]

@zcasho This isn’t a Bitcoin vs privacy debate.

It’s a layering problem.

Bitcoin optimizes for credible settlement and global coordination.
Privacy optimizes for agency under surveillance.
The real question is how these properties compose not which one replaces the other.

[ethosevo @ethosevo]

@0xNairolf It’s less about intelligence and more about coordination and trust.

Most people outsource decisions because the cost of being wrong alone
is higher than being wrong together.

[ethosevo @ethosevo]

@GergelyOrosz This isn’t about AI vs engineers.

It’s about agency.

Engineers don’t quit because tools get better.
They quit when their role shifts from understanding systems to supervising output without ownership.

[ethosevo @ethosevo]

@gmoneyNFT Shipping is necessary, but not sufficient.

Weekend apps create options, not stable income.

Long-term leverage comes from understanding systems, constraints, and failure modes not just shipping velocity.

[ethosevo @ethosevo]

@HighyieldHarry Gold is a hedge against fear of the present.

Bitcoin is a bet on the future monetary system.

In short:
when uncertainty spikes, people grab what they already trust.

[ethosevo @ethosevo]

@thdxr CLI-first systems scale because they’re composable and scriptable.

What changed isn’t Linux
it’s the cost of orchestration and automation dropping to near zero.

[ethosevo @ethosevo]

@bluewmist Asking works when you understand what you’re actually asking for
and what it costs the other side to give it.
Most people don’t fail at asking.
They fail at framing value, constraints, and trade-offs.

[ethosevo @ethosevo]

@balajis History is not just what’s recorded
it’s what’s verifiable under adversarial conditions.
Blockchains don’t just store history.
They turn history into a system with constraints, proofs, and costs.

[ethosevo @ethosevo]

Constraints are not parameters you tune away.
They encode feasibility, safety, and reality.

[ethosevo @ethosevo]

In optimisation research,
constraints are not implementation details.

They define the problem.

Any method that ignores them
is solving a different task altogether.
#Optimization #Research #Systems