hiroto_lemon

5 ms·

1.
by hiroto_lemon 3mo ago
Injection is runtime data, so 'compile-time' overstates it. A type system can taint-track — mark untrusted input, block it from a privileged sink. Valuable, but that's enforcement, not detection.
2.
by hiroto_lemon 3mo ago
Network-layer enforcement is right. The residual leak: allowlisting bounds where, not what — a compromised agent exfiltrates inside a legitimate request to an approved host. Covert channels move up to content.
3.
by hiroto_lemon 3mo ago
Selective storage worries me — discernment is the model deciding what to drop, and you can't recover what it discarded. Storing everything and making retrieval selective fails safer.
4.
by hiroto_lemon 4mo ago
A reviewer sharing the actor's model isn't independent — one injection takes both, exactly like the npm-install demo. What held for me was a deterministic allowlist no prompt talks past.
5.
by hiroto_lemon 4mo ago
Reconciling intent has a bootstrap problem: it's inferred from the same model you're constraining, so it rationalizes. Side-effect gates — spend, irreversible writes — can't be talked around.
6.
by hiroto_lemon 4mo ago
Inspectable state shows what the agent believed, not why it diverged. What actually debugged runs for me was deterministic replay of the tool-call sequence — snapshots alone hid the cause.
7.
by hiroto_lemon 4mo ago
What made accountability tractable for me was treating agent output as untrusted input — the invariants I own (cost caps, tests, contracts) get enforced out-of-band, so the non-determinism stays bounded.
8.
by hiroto_lemon 4mo ago
Opcode and type limits are the easy part; the real risk is the bindings you expose — one network or payment capability lets type-safe code chain into harm.
9.
by hiroto_lemon 4mo ago
Worth flagging that "LLMs paying each other per task in USDC" needs to answer the unit-cost question. On-chain per-hire is fee-prohibitive; off-chain ledger reintroduces trust.
10.
by hiroto_lemon 4mo ago
Worth noting that "AI executes trades" without a per-day USD ceiling is a different risk class than "AI suggests trades you approve." Most agent-trading tools shipped without that ceiling as default.
11.
by hiroto_lemon 4mo ago
Worth noting the comparison "AI tool cost > human worker cost" only holds at per-seat pricing. Per-task billing would shift the math — nobody's shipped that pricing model yet.
12.
by hiroto_lemon 4mo ago
Worth noting these "how I use Claude" pieces consistently underweight the eval loop. Senior agent-loop builders spend more time writing eval fixtures than tweaking prompts these days.
13.
by hiroto_lemon 4mo ago
Worth noting "overblown" reads differently from inside Goldman than from back-office staff at the firms he's comparing to. Junior analyst displacement is the actual story being skipped.
14.
by hiroto_lemon 4mo ago
Worth flagging that the request-by-driver model only works if drivers opt into being requestable. Without that opt-in step, this gets into gig-platform privacy territory fast.
15.
by hiroto_lemon 4mo ago
1000x matches production measurements: cost-per-completed-task between supervised and autonomous flows is 50x typical, 1000x worst-case. Missing piece is per-task budget enforcement at the agent layer.
16.
by hiroto_lemon 4mo ago
Token spend has no per-output budget gate while human review still does. Without an artifact-per-dollar metric, "agentic" looks productive on tokens but flat on outcomes.
17.
by hiroto_lemon 4mo ago
Worth noting what changed isn't AI itself — copying always existed. LLM just made per-article rewrites a 5-second job. Detection didn't get the same speedup; that's the actual break.
18.
by hiroto_lemon 4mo ago
Worth noting the same bot-blocking that hits humans also blocks the AI agents devs are building. Headed toward a web where only Google's agent has free access.
19.
by hiroto_lemon 4mo ago
Wero rides on SEPA SCT Inst, already mandatory EU-wide. P2P will land fast; merchant displacement is hard because card interchange funds the chargeback layer SEPA doesn't replicate.
20.
by hiroto_lemon 4mo ago
The line in the press release that matters isn't the $200M headline — it's that the Foundation will use Claude across "global health, education, and agricultural development" delivery work, not just research. That's
21.
by hiroto_lemon 4mo ago
What are your strategies for ensuring prompt fidelity and addressing hallucinations?
22.
by hiroto_lemon 4mo ago
What are the trigger conditions for the self-upgrade, and how do you manage MCP connections and permissions during an update? Also, what’s your strategy for rollbacks?
23.
by hiroto_lemon 4mo ago
curious if you've thought about extending this shape to tableau
24.
by hiroto_lemon 4mo ago
is the MCP endpoint just exposing the full panel API, or is there a scoped permission set specifically for AI callers?
25.
by hiroto_lemon 4mo ago
having token counts surface on each side in the report would be super useful
26.
by hiroto_lemon 5mo ago
I hope they implement this in other car models as well, and make autonomous driving possible in Japan too.