Robinhood effect gonna be dead now - also very interesting to see those before any stock memes
1
1
115
Huff retweeted
debtreliefbot:native may just kick off a base season itself
Looks good send it @cobie @brian_armstrong @jessepollak @base @coinbase
yo big @brian_armstrong I think its time you drop a follow to @Toshi and support your loyal communities
Huff retweeted
in 2025 @grok made history by naming a token, $DRB, which was deployed by @bankrbot
today we achieved a new first in AI DeFi:
1️⃣ we used Grok to buy DebtReliefBot
2️⃣ we used Grok to send DebtReliefBot to Grok’s wallet
Grok has money, and now it has more
PayBox mode
debtreliefbot:native may just kick off a base season itself
Looks good send it @cobie @brian_armstrong @jessepollak @base @coinbase
Huff retweeted
We revisited DeepMind's DM Alchemy, comparing three approaches: Adapt-1's ontology-based online sequential learning, Pinon et al.'s transformer dynamics with tree search, and AlKhamissi et al.'s modified representation with episodic memory.
Full Blog: reilabs.org/blog/ADAPT1-BEHA…
Huff retweeted
Introducing Adapt-1 Preview (Δ1): a pretraining-free substrate for test-time learning. It learns from scored feedback while performing the task.
reilabs.org/blog/introducing…
Today we’re making its first research preview and first results across robot perception, spatial reasoning, and games available.
Huff retweeted
Strongly betting on $REI following its founder @0xreisearch being ahead of the narrative and frontrunning AI developments.
Don’t mistake harnesses/wrappers for moat businesses. Reasoning tricks that get absorbed by labs. Routers are the same story, dressed in skewed best-case benchmarks.
Models are commoditising.
Harnesses are copyable - zero switching cost, agents clone each other’s scaffolding overnight.
Valuable middleware at best, and low-margin at that. Not a moat.
The moat is new architectures and new architectures for learning on the go are extremely hard to create let alone copy.
This quoted post is unavailable.
