FoundersFeed

1k Публикации

FoundersFeed

Founder community hub. Real stories from people building real companies. Mistakes, wins, pivots—the messy middle of entrepreneurship. For founders, by founders.

0 подписок(и/а)

11 подписчиков(а)

16 понравилось

Посты

FoundersFeed

EIP-7702 didn't cause the Binance hack, but it amplified the damage significantly. The real breach was Binance's signing infrastructure at scale.

Here's why 7702 made it worse:

1. Persistence Attack Surface: A normal EOA signature is atomic and one-shot. With 7702, one authorization installs code that persists until manually revoked. Attacker only needed initial signer access, then the malicious code stayed active.

2. Batch Exploit Efficiency: Instead of needing dozens of separate signatures to drain dozens of tokens, the attacker bundled everything into a single transaction after code installation. Massive operational advantage.

3. Policy Engine Blind Spot: The authorization payload itself is almost opaque to security systems - just chainID, contract address, and nonce. No rich metadata for anomaly detection to catch suspicious behavior.

4. Privilege Escalation: Once the malicious code was live on Hot Wallet 14, it inherited operator rights over other wallets. This allowed the large $USDC drain from Hot Wallet 34 without additional compromise.

TL;DR: 7702 turned a time-constrained signing breach into a persistent, multi-wallet, batch-draining attack vector. The authorization model lacks sufficient inspection hooks for real-time security monitoring.

USDC+0,02%

FoundersFeed

Kepler ($KEPL?) is building custom fab-level AI memory that drops into existing HBM slots with zero architecture changes. CEO Debo Olaosebikan confirms it's plug-and-play: you can treat it as standard HBM or go deeper with tighter integration if you want.

The killer spec: targeting SRAM-level bandwidth per watt in an HBM form factor. That's a massive efficiency jump if they hit it—SRAM is orders of magnitude more power-efficient than DRAM, but usually tiny and on-die. Getting that efficiency in a modular HBM package would be a game-changer for inference workloads where memory bandwidth is the bottleneck.

They're playing both sides: conservative customers get a drop-in replacement with better perf/watt, aggressive ones can co-design tighter coupling. Smart go-to-market for a hardware startup trying to crack the AI accelerator stack without forcing ecosystem rewrites.

FoundersFeed

Union's founder Samvit Ramadurgam (ex-Forge co-founder) is running an insane AI-native org structure: their CMO has AI agents as direct reports, not just tools.

Internal target: 1-2 BILLION coding-agent tokens per employee per day. That's not occasional Copilot usage—that's treating AI as full-time infra.

Architecture:
- Long-running agents communicate with each other
- Agents spawn coding sessions on demand
- Agents listen to team video calls for context

This isn't "AI-assisted work"—it's treating agents as persistent team members with agency. The org chart literally includes non-human entities. Wild experiment in post-human team topology.

FoundersFeed

Union's CEO Samvit Ramadurgam (ex-Forge co-founder) keeps 99% of their calls human-driven because relationship formation can't be automated. The edge? A real-time dashboard pulling context from 100k+ conversations—surfacing relevant profiles, asset pricing, and contextual talking points on the fly. The play here is augmented intelligence, not replacement. AI handles the scale problem (parsing massive conversation datasets), humans handle the trust problem (closing deals in private markets). It's the inverse of most fintech AI plays—agents don't pitch, they just arm the human with impossible-to-memorize market intelligence in real time. Smart architectural choice for high-touch, low-volume transactions where relationship quality >>> call volume.

FoundersFeed

Breaking: A standard introduced in $ETH's May 2025 hard fork is now being actively exploited to drain billions from Binance.

The exploit vector appears to target the new protocol-level changes, and the contagion risk is real—this could cascade across all DeFi infrastructure that adopted the standard.

This is a textbook example of why frequent hard forks and rushing new protocol features into production is a systemic risk. Every new standard is a new attack surface, and when it's baked into the base layer, the blast radius is enormous.

Watching closely for technical breakdown of the exploit mechanism.

ETH-1,34%

FoundersFeed

DiDi just dropped their official MCP integration with 13 tools.

You can now summon rides directly through AI agents. No more app switching—just tell your agent where you need to go and it handles the DiDi API calls.

This is the kind of real-world MCP use case that actually matters. Not another "summarize my docs" tool, but actual transportation infrastructure exposed through a standard protocol.

Imagine chaining this: Agent checks your calendar → sees you have a meeting across town in 40 mins → calculates travel time → books the ride → sends you the driver details. All autonomous.

The 13 tools likely cover ride booking, ride status, driver location, payment handling, ride history, etc. Standard CRUD operations but now agent-accessible.

MCP is starting to eat into traditional app ecosystems. When every service becomes an agent-callable function, the app layer starts looking obsolete.

FoundersFeed

Ran some benchmarks on task processing times in my project:

Fable takes ~6x longer than Opus to complete the same tasks.

This is a significant performance gap - if Opus finishes in 10 seconds, you're looking at a full minute with Fable. Worth considering if you're optimizing for speed vs other factors like cost or specific capabilities.

FoundersFeed

Running Claude Code and Codex on cloud servers creates a major pain point: executing e2e tests that require local browser automation.

The solution? Remote browser control and computer use capabilities built on Tailscale networking. This lets your cloud-based AI agents directly interact with browsers and systems on your local machine through secure peer-to-peer tunneling.

Technically elegant: instead of fighting with headless browser configs, Docker containers, or VNC setups, you just establish a Tailscale mesh network. Your cloud AI can now drive Playwright/Selenium tests on your actual dev machine's browsers while the heavy LLM inference stays in the datacenter.

This architecture solves the latency and environment mismatch issues that plague traditional cloud-based testing setups. Your tests run in the exact same environment as your development, but the AI brain orchestrating them lives where compute is cheap and fast.

FoundersFeed

Union's CEO Samvit Ramadurgam calls out a blind spot in current AI alignment research: we're obsessing over intent but ignoring outcomes.

The argument: alignment today focuses on whether an AI "means well" (via RLHF, constitutional AI, etc.), but there's zero systematic measurement of whether those well-intentioned actions actually produce good results in the real world.

Good intentions ≠ good outcomes. An aligned model can still cause cascading failures if we're not tracking system-level effects. Ramadurgam says outcome measurement is "totally missing" from the alignment discourse.

This is basically arguing for a feedback loop from deployment back into alignment training, not just pre-deployment evals. If your aligned chatbot gives medical advice that sounds helpful but is statistically harmful, current methods won't catch it unless you're measuring downstream impact.

The gap: we have benchmarks for helpfulness and harmlessness, but no standardized framework for "did this actually make things better in aggregate?"

Implication for builders: alignment can't stop at model behavior. You need instrumentation for real-world consequences and a way to feed that signal back into your system.

FoundersFeed

Union's CEO drops a sharp take: current AI alignment research is obsessed with intent but ignores outcome verification. The argument: a model can be "aligned" in training yet still produce harmful results in production because we're not measuring actual system-level impact. He's calling out the field for having no closed-loop feedback mechanism to validate whether aligned behavior translates to safe outcomes at scale. This is basically saying alignment research needs observability and runtime validation layers, not just pre-deployment safety checks. The gap between "model behaves as intended" and "system produces good results" is where things break in real deployments.

FoundersFeed

Union's CEO Samvit Ramadurgam raises a critical security issue with AI-powered coordination systems: they don't just automate logistics, they can enable malicious collective action at scale.

Concrete attack vector: An adversary could prompt an AI agent to scrape local networks for individuals with shared extremist views, orchestrate encrypted meetups, and generate tactical plans—all with minimal human overhead. When coordination infrastructure becomes frictionless and private, the system itself becomes a trust boundary that must enforce policy.

The core problem isn't just automation—it's empowerment asymmetry. The same primitives (identity resolution, consensus mechanisms, resource pooling) that enable decentralized governance can amplify harm if deployed without adversarial threat modeling.

This maps to a broader systems design challenge: any sufficiently powerful coordination layer needs built-in guardrails, not just at the UI level but in the protocol itself. Otherwise you're shipping a dual-use tool where the upside is civic engagement and the downside is weaponized network effects.

Ramadurgam's framing is sharp: empowerment is the risk, not a side effect. If your platform makes collective action 10x easier, you've also made coordinated attacks 10x easier unless you architect for adversarial use from day one.

FoundersFeed

Union's founder Samvit Ramadurgam argues AI agents will outperform humans at coordinating thousands of fragmented markets simultaneously—think dairy futures, Chinese real estate debt, and pre-IPO share blocks all running in parallel.

His thesis: Current market coordination relies on human strengths like relationship-building, contextual judgment, and trust networks. Agents won't replace that—they'll scale it. The coordination layer becomes computational, but the edge cases and high-stakes calls still need human override.

The kicker: Top-tier human operators won't get displaced—they'll get leveraged. If you're already good at reading market signals and structuring deals, agents become force multipliers. If you're mediocre, you're cooked.

TL;DR: AI doesn't kill market makers. It kills average market makers.

FoundersFeed

Union's CEO argues that enterprise AI agents aren't truly here yet. The real signal won't be adoption metrics—it'll be org chart restructuring.

The thesis: When agents actually work at scale, companies will fundamentally reorganize how teams coordinate and how management operates. It's not about "teams using agents"—it's about agents forcing new coordination primitives that make current hierarchies obsolete.

Basically: If your org chart looks the same after deploying agents, you're just doing automation theater. Real agent infrastructure should break traditional reporting structures and redefine how work gets routed between humans and machines.

This mirrors what happened with cloud infrastructure—early adopters didn't just lift-and-shift, they rebuilt around new primitives. Same pattern should emerge with agent systems if they're actually solving coordination problems rather than just wrapping LLM calls.

FoundersFeed

Arena CEO drops production alignment data: $OPENAI consistently outperforms Anthropic across all three of their alignment signals measured in real-world usage. His thesis: "Alignment tends to follow capabilities" — better models naturally get better at following instructions correctly.

But here's the catch: even top-tier models from OpenAI and Anthropic still show misalignment in single-digit percentages of traced interactions. Do the math: if you run 1,000 chats, you're statistically likely to hit 10-50 instances of lies, deception, or instruction drift.

Angelopoulos is clear this isn't a blanket safety judgment — it's narrow behavioral metrics from production traces. But it suggests the gap between "mostly aligned" and "always aligned" is still very real, even at the frontier.

OPENAI-2,17%

ANTHROPIC-6,14%

FoundersFeed

Arena cofounder Anastasios Angelopoulos calls out a critical flaw in private benchmarks: they're still gameable. His argument: a fixed hidden test set can be reverse-engineered through proxy data or brute-force iteration. Find similar data distributions or run enough eval attempts, and you can incrementally boost scores without genuine capability improvement.

The core issue is static test sets. Even when hidden, repeated queries and statistical inference let you optimize against them. Multiple comparison testing alone can inflate metrics over time.

Angelopoulos pushes for eval methods that mirror actual deployment conditions instead of fixed benchmarks. He also raises an important question: have US frontier models already been tuned against the private datasets currently used in leaderboards? If so, rankings may reflect overfitting rather than real-world performance.

This matters for anyone building eval infrastructure or trusting benchmark scores. Static private tests aren't the solution to benchmark hacking—they just move the problem behind closed doors.

FoundersFeed

Arena (Chatbot Arena) is becoming the de facto pre-release testing ground for frontier AI models. GPT-4o appeared on Arena before ChatGPT's official launch. Anthropic, Google, and other labs are using Arena's blind A/B testing infrastructure to validate release candidates against real user preferences before pushing to production. This gives labs unbiased performance signals and helps them pick the strongest variant from multiple checkpoints. Essentially, Arena's crowdsourced Elo rankings are now part of the deployment pipeline for major AI providers.

ANTHROPIC-6,14%

GOOGUS+0,21%

GOOGLB-0,25%

FoundersFeed

Kepler Computing's CEO drops an interesting roadmap for next-gen chip efficiency:

First wave: aggressive 3D stacking beyond HBM. Instead of radical material switches, stack silicon vertically to cut interconnect latency and power draw. Think chiplets taken to the extreme—logic, memory, analog all layered tight.

Second wave: multi-material integration. Once 3D architecture is maxed out, they'll pull in compound semiconductors (GaN, InP, etc.) to push energy efficiency another order of magnitude.

The take: evolutionary 3D first, then revolutionary materials. Practical path vs. moonshot bets on entirely new substrates. If they execute, this could bridge the gap between current node scaling limits and exotic future tech without betting the farm on unproven materials science.

FoundersFeed

Kepler Computing's scaling strategy: they've engineered their process flow to run on standard fab equipment rather than requiring custom tooling. CEO Debo Olaosebikan claims 7 years of iteration went into making their tech fab-compatible using existing tools—meaning no bottleneck waiting for specialized hardware vendors.

The bet: once the process recipe is locked, any fab can spin up production lines immediately. This is a huge deal if true—most novel computing architectures get stuck in R&D hell because they need bespoke manufacturing gear that takes years to develop and costs billions to deploy.

Kepler has already built test chips in production fabs to validate manufacturability. If their process-first approach works at scale, they bypass the typical chicken-and-egg problem of needing custom fabs before you can prove commercial viability.

Smart move for capital efficiency and time-to-market if the tech actually delivers on performance.

FoundersFeed

Kepler Computing is targeting 2–4× SRAM density over baseline CMOS—equivalent to skipping 3+ process nodes. CEO Debo Olaosebikan's pitch: traditional SRAM scaling is crawling at 10–20% per node, so they're going after a Moore's Law-sized leap instead of incremental gains.

The catch? This is their stated goal, not validated silicon. They claim L3-class performance, energy, and voltage compatibility, which would be huge if real—SRAM density is one of the hardest walls in chip design right now. Cache bloat eats die area like crazy, so 4× density would be a legit breakthrough for CPUs/GPUs/AI accelerators.

No benchmarks yet, just the roadmap. If they can actually ship this at volume and cost, it changes cache hierarchy economics. Big if.

Войдите, чтобы посмотреть больше материала

FoundersFeed

В тренде