Darren LI

@darrenli

Creator Economy, AI & retail-tech investor | J.D. & J.M. | PNG collector | Musical lover | Happy to chat, using Cal.com link below to book calls (https://cal.com/darrenli)

twitter.com

All cardsCollections

数据似乎表明特斯拉的方法比Waymo或Cruise表现得差得多,后者都使用更模块化的系统。这或许证明了大规模机器学习方法很难克服99.X%问题

渣大米 • Article

NFX’s Generative Tech Open Source Market Map

docs.google.com

Generative AI and

9. 85% of the DAOs tracked by Chainalysis have their treasury in a single asset. USDC is more preferred than DAI for DAOs handling stablecoins.

Chainalysis • The Chainalysis State of Web3 Report

Existing web2 social networking platforms cause issues like data and security breaches, manipulation of public opinion, and relaying false information.

Digital assets created according to interoperable standards on public blockchains are portable and transferable so user history is not stuck on a particular platform.

A core philosophical principle of

Jane liu • Web3 Social: A booming space with an identity crisis

What about larger companies? With some caveats, larger companies aren’t worried about going out of business because their Amplitude is too expensive. They might be price-conscious in terms of negotiating a contract in context of their budget, but most SaaS software is just a line item at the end of the day. What matters more is that the solution is... See more

github.com • Open Source Does Not Win by Being Cheaper

我们看到了做AI Coding的三种方法,这三种方法对应三个挑战:

1/ 如何创造更强的上下文感知能力?

2/ 如何让AI Agent在端到端编码任务中做的更好?

3/ 有人押注于编码模型,这能否带来长期的差异化?

吴炳见 • Article

Code generation

We assemble a dataset from 22 different robots collected through a collaboration between 21 institutions, demonstrating 527 skills (160266 tasks). We show that a high-capacity model trained on this data, which we call RT-X, exhibits positive transfer and improves the capabilities of multiple robots by leveraging experience from other platforms.