Markets are starting to simmer, so if you're not locked in right now I'm not sure wyd. Few projects below that I think will go higher, last one undervalued by a factor of 100 $SURPLUS >dev @mac_eth formerly of Coinbase, CashApp >recent s/o from Head of Product at Venice >4.4M FDV The size of the market they're attacking is significant, with AI compute spending at roughly a ~150B sized market today predictions of this moving towards 250-500B by 2030. Currently the landscape is human -> AI, whereas in future this'll be human -> 100 agents -> thousands of AI calls. Surplus is interesting bc it turns unused AI consumption rights into tradeable inventory. Think how airlines hate flying with empty seats because each unfilled seat is lost revenue that can't be recovered. AI consumption rights have the same problem: if you've pre-purchased capacity but aren't using it, that compute time disappears. Thanks to Surplus Intelligence, those AI "consumption rights" don't have to sit idle anymore. The tldr is: if compute becomes the oil of AI, Surplus could become the energy grid for the agent economy. You're buying exposure to a new liquidity layer forming around AI consumption itself where unused AI capacity becomes a tradeable inventory... burn mechanism likely moving forward, no ceilings here IMO $GRANTR >dev @alexandrechopan formerly of Metamask, Consensys, Trust Wallet >1.6M FDV Alexandre is working on EIP-8141 as touched upon by Vitalik recently (x.com/VitalikButerin/status/…) and the only current exposure is @grantr_id. Alexandre is literally someone who helped power ERC-4337 adoption now building its native replacement. Token wasn't launched by him, but he's publicly claiming contract fees onchain and it's still ~1M market cap. If EIP-8141 becomes the future many expect, this looks like one of the only real exposure plays before the narrative wakes up. IMO the interesting part is protocol upgrades rarely capture all the upside/ value themselves, more often than not, the interfaces/apps/ecosystems built around them are often where the outsized upside ends up accruing. $INSTACLAW >doxed founder and advised by a former Anthropic/Google employee >1.2M FDV OpenClaw provides the open-source agent framework/runtime, while InstaClaw removes the friction of using OpenClaw: provisioning servers, keeping agents online, connecting Telegram/Discord/WhatsApp, persistent memory, skills, uptime monitoring, and VM management. Most people won’t configure workflows or self host servers, Instaclaw removes this friction. Put simply, they're the front-end, or Apple iPhone experience for OpenClaw. They'll be acting as the agent infrastructure for @joinedgecity upcoming agent-human intel experiment, which aims to be the largest ever of its kind. This is a month long experiment where builders test new tech like InstaClaw. Edge City is affiliated with the likes of World Foundation, Uniswap Foundation, Stanford Blockchain, etc. $SPETTRO >identified by @ycombinator as a top candidate for their start-up school >2nd place in @cognition recent hackathon, a company valued at $25B >Play Store app in the works >80K FDV (absurd) @cesp2099 is a cracked dev that's won multiple hackathons and has traction on GitHub. Spettro is a AI-native system that can actually execute and coordinate software creation end-to-end. Spettro isn't another AI wrapper where you slap a chatbot onto a code editor and call it innovation. The interesting part is the orchestration layer itself. Multiple specialized agents can coordinate, delegate tasks, preserve context and execute work in parallel like an autonomous software team. Proving the point, they recently ported an entire application from Go and C to Kotlin and C in under 30 minutes for roughly $5 in AI credits. To put that in perspective, this would take days from snr engineers and hundreds if not thousands of dollars... Essentially, this is functioning autonomous AI infra handling architecture translation, Android migration, bug fixes, modern APIs and release generation _______________________ IMO these are the best candidates to pull a GITLAWB type of move, and once more projects consistently crack 10M, then 100M... the floors on all of these will be significantly higher. Still absolute madness that SPETTRO is sitting at 80K, are solbros ok? [prob not]