Grok
From Mecha Hitler to SOTA rare-disease diagnosis in children? @spacexai's @grok 4.6 has taken the 👑 on RareBench, edging out @AnthropicAI Claude Opus 5 for about 1/3 the cost. This was not on my 2026 bingo card!
@deepseek_ai's new v4-pro-0813 model underperformed my expectations, v4-flash, and seemingly the entire internet's. We accessed using DeepSeek's 1P API on the day of release and I almost wonder if they didn't switch over their model endpoint correctly. We will re-benchmark and report back.
@Zai_org has attracted a following with GLM5.2, but they, too, underperformed. This doesn't surprise me because when I compared GLM and @Kimi_Moonshot K3 for coding use-cases, I found Kimi substantially stronger, but the internet seems to love this model.
574
729
3,544
2,356,193
Crypto is entering a phase where risk, privacy, distribution and real-world utility are starting to matter more than narratives alone.
That shift is showing up across completely different parts of the market.
Take HLP.
A 15–35% APY looks attractive until you consider where that yield comes from.
HLP takes the other side of traders, which means unusual market behavior can become LP losses. The $13.5M JELLY incident showed how quickly that risk can materialize.
The lesson is bigger than HLP:
Yield is only as good as the risk sitting underneath it.
The same idea applies to privacy.
Public blockchains give us transparency, but serious financial activity also needs confidentiality.
That’s why projects like @Zcash, @AleoHQ, @aztecnetwork, @namada, @OasisProtocol, @fhenix and @nillion are building different approaches to private computation, transactions and data.
The goal isn’t necessarily to hide everything.
It’s to prove what matters without exposing everything else.
Then look at new chains.
Robinhood Chain shows how distribution and speculation can create economic activity quickly, while launchpads capture some of the earliest fees.
That makes the emerging launchpad ecosystem around Arc interesting too, with platforms such as @actfunxyz, @arcjetfun, @Chaingpt_Pad, @SharcFun and others competing for that asset-creation layer.
But speculation is only the beginning.
The bigger opportunity is what happens when those networks mature into stablecoins, lending, RWAs and payments.
And that brings us to physical AI.
Robots are moving from controlled industrial environments toward autonomous systems that can perceive, reason and act.
For that economy to work at scale, machines need more than AI.
They need identity, positioning, data, coordination and payments.
That’s where projects like @peaq, @GEODNET, @Auki, @opentensor, @FabricFND, @PrismaXai, @konnex_world and @xmaquina become interesting.
They are attacking different pieces of the same emerging machine economy.
Even institutional markets are moving in this direction.
The withdrawal of the Cardano, Hedera and Polkadot ETF filings is a reminder that institutional access isn’t simply about market cap.
Liquidity, demand, market depth and usable infrastructure matter.
So when I look across HLP, privacy protocols, new L1s, launchpads, robotics and ETFs, I don’t see completely separate narratives.
I see one broader transition:
Crypto is becoming more focused on real economic infrastructure.
The market is slowly moving from asking:
“Which narrative gets attention?”
to:
“Which system can actually handle capital, risk, users and real-world activity?”
✍️ Conclusion:
The next cycle may reward infrastructure more than hype.
Not because narratives disappear, but because the projects that survive will need something stronger underneath them:
real users, real activity, real cash flows and infrastructure people actually need.
Aug 13, 2026 · 7:30 PM UTC
45





