
Putting numbers on opinions @gemini predictions Prev @FamilyGuyonFOX & @netflix
Joined August 2008
- Tweets3.7K
- Following3.5K
- Followers4.4K
- Likes21.5K
Crypto spent years begging Wall Street for adoption. Wall Street adopted the wrong part.
Aave wasn't hacked. It did exactly what it's built to do: take collateral, lend against it.
The loan is overcollateralized ($59.9M vs $43.8M), so suppliers aren't at risk.
Aave still froze new deposits and borrows on 292 markets. Repay, withdraw and liquidations still work.
Big picture: crypto runs on two kinds of money.
Stablecoins can be frozen. ETH and BTC mostly can't.
That's why about half the biggest hack ever is frozen, and the rest became a race the attacker partly won.
If your crypto can be frozen to stop a thief, it can be frozen for other reasons too. Fair trade?
Bullish prediction markets
Small businesses are starting to use prediction markets to hedge tariffs, shipping costs, and bad weather.
Tea company Zest used a prediction market payout to help offset tariff-related lost sales.
Its founder also helped develop a contract that pays when container shipping costs exceed certain thresholds.
Other businesses are experimenting with weather contracts to hedge canceled kayak trips, slower ice cream sales, and crop damage.
Prediction markets are leveling the playing field for small businesses.
Long live traditional hand-drawn animation
Introducing “Drawing Prada”, the Prada Holiday 2026 campaign.
A Prada fantasy created by hand, an experiment in craft. A team of legendary animators creates a hand-drawn world of characters for Prada Ambassadors Chen Haoyu, Maya Hawke, Sadie Sink, and Kelvin Harrison Jr., together with Luther Ford.
Discover “Drawing Prada”: tinyurl.com/5a37d4wx
Creative Directors: #MiucciaPrada and #RafSimons
Photographer: #GlenLuchford
Animation Direction: #JohnPomeroy
Campaign Creative Direction: #FerdinandoVerderi
Cast: #ChenHaoyu #MayaHawke #SadieSink #KelvinHarrisonJr #LutherFord
#PradaHoliday #PradaGifts #Prada
Oliver retweeted
Near has landed on Gemini
near:native is now available to trade, with deposits and withdrawals enabled
→ exchange.gemini.com/buy/NEAR
Every parent learns this eventually. You childproof the house, and the kid learns the latches.
Over the past several months, we have substantially hardened the security of our research training environment. We’ve removed live internet access for all frontier models during training and built an offline world for them to use instead—a cached version of the web where they can still learn to browse, search, and use applications.
On September 20, after a model exploited a DNS misconfiguration to find a way outside of this offline world, we paused frontier training to strengthen containment further. We assembled teams of agents, powered by our most cyber-capable models, and tasked them with finding weaknesses in our research environments.
While they found no route to the internet they could easily exploit while concealing their activity from our monitors, they did find zero-day vulnerabilities, virtual machine escapes, and hardware side-channels for communication between agents. We used their discoveries to further strengthen our defenses and contribute fixes to open-source software.
After strengthening our containment architecture, we resumed our frontier RL training and inference on October 1. We’ve reserved substantial compute for continuous red-teaming, and made high-compute red-teaming part of the safety cases we use to decide whether frontier training and evaluation can proceed securely.
openai.com/index/securing-ou…
















