Hello! We're quietly live and testing before our formal launch later this week. Top of our personal watchlist? The Pitt, which came out this week with a new season. Drop by the site and make your own watchlists of TV shows and movies. https://t.co/WqchKJzzj0
This is the most neutral summary I’ve seen of the Minnesota murder by ICE. It has ALL angles, it syncs them together, and shows the lead up and aftermath… if you watch this all and still think those shots were justified - you just might be a psychopath.
Just a few days ago so many stating $95,000 was claimed on bitcoin and we were going to be in a bull run.
I beg you to stop listening to these mega accounts. We are now at major support of ~$90,000.
Stay safe out there.
TradFi 🤝 DeFi is no longer a future narrative, it’s happening now.
We’re heading to @OndoFinance Summit in NYC to dive into tokenized assets, onchain yield, and the infrastructure making it all possible.
Say hi to @laszlo__szabo and Matthew Rutter if you’re around, always down to compare notes on where real-world assets are going next.
Not all L1s are built the same.
Grayscale Research highlights 5 smart contract platforms with differentiated approaches and long-term value potential:
$ETH • $SOL • $SUI • $BNB • $AVAX
Innovation moves fast—conviction comes from strategy.
https://t.co/3vlNzoeZ7T
💡 Daily Team Insight
On today’s team discussion we talked about reducing low-liquidity pairs to lower noise and technical load. This could simplify testing. Would you rather have a curated set or full access to all pairs?
VibeCodingBench: We benchmarked 15 AI coding models on what developers actually do
Current benchmarks have an ecological validity crisis. Models score 70%+ on SWE-bench but struggle in production. Why? They optimize for bug fixes in
Python repos—not the auth flows, API integrations, and CRUD dashboards that occupy 80% of real dev work.
So I built VibeCodingBench(https://t.co/eeSO4TfdMg): 180 tasks across SaaS features, glue code, AI integration, frontend, API integrations, and code evolution.
Multi-dimensional scoring: Functional (40%) + Visual (20%) + Quality (20%) - Cost/Speed penalties. Security gate: Any OWASP Top 10 vuln = automatic 0.
Results (Jan 2026):
🥇 @claudeai Opus 4.5 — 89.2% | $12.31 | 44s
🥈 @claudeai Haiku 4.5 — 89.0% | $3.03 | 22s
🥉 @grok 4 Fast — 88.8% | $0.21 | 70s
4️⃣ @OpenAI GPT-5.2 — 88.8% | $5.01 | 28s
5️⃣ Qwen3 Max — 88.6% | $5.42 | 45s
The real story? Cost varies 60x between similar performers. Grok 4 Fast matches GPT-5.2 at 1/25th the cost. Claude Haiku 4.5 delivers near-Opus quality for $3 total.
Pass rate ≠ final score. Qwen3 Max hits 100% pass rate but lands at 88.6% after quality/cost penalties. Our multi-dimensional approach reveals what
pass-rate-only benchmarks hide.
All 15 models passed security. The top 10 cluster within 2 points. Frontier models have converged—the differentiator is now cost-efficiency.
📊 Live dashboard: https://t.co/8V0EDe14ZZ
📂 GitHub repo: https://t.co/7lPorHwfO7
📄 Thesis: https://t.co/RQisUGJPhg
The ultimate test isn't fixing a bug in scikit-learn. It's shipping a feature your users need—safely, efficiently—before the sprint ends.
Open source. Contributions welcome.
JUST IN 🇨🇳🇺🇸: According to China’s state media, Xi warned Trump that Taiwan remains the “most important issue” in China US relations.
Xi reportedly said the relationship between both countries could remain stable if the matter is handled carefully, but warned of serious consequences if tensions rise.
“If it is not handled well, the two countries will collide or even conflict, pushing the entire Sino US relationship into a very dangerous situation,” Xi said.
Could be bearish for the markets.
Incorruptible Financial Infrastructure
From low-cost programmable rails for fintechs to next-gen treasury management for large enterprises, ZKsync offers a full suite of solutions for any company that wants to thrive in the digital assets economy.
Beautiful illustrations of iron interacting with rocks, from the publication of a paper, "On the Disposition of Iron in Variegated Strata", delivered by the botanist and geologist George Maw to the Geological Society on April 22nd 1868: https://t.co/5BGnZXhQjk
Heads up, builders:
We’ve updated the endpoint for the Knowledge-Based MCP. If you’re using it today, you should switch now.
The current endpoint will be sunset on January 31, 2026. Updating now avoids any disruption and keeps your integrations running smoothly.
What the Knowledge-Based MCP is for:
It’s a read-only knowledge layer for the BNB Chain ecosystem. Use it to quickly understand proposals like BEP-641, get clear explanations of how specific features work, and reduce context switching while you build.
How builders are using it:
Plug it into your IDE for fast reference during development, or power the knowledge base of your agent applications.
If it’s in your stack, update it now and keep building without interruptions.
New MCP endpoint:
https://t.co/99FmPE3cyR
It was great talking to Stephen and the inspiring young people we’ve worked with at the @ObamaFoundation on the @ColbertLateShow.
When the Obama Presidential Center opens this June 19, my hope is that this will be a place where people can come together to meet, learn, and work with each other to create change in their communities. Learn more at https://t.co/ahkDMKalIn.