HN user

yonz

258 karma

Founder Vella.ai, previously at LinkedIn and Qualcomm, Harvard MBA

[ my public key: https://keybase.io/yonz; my proof: https://keybase.io/yonz/sigs/Nz2P_8G1t8p_n8V3eTNFn6jtO3jzGLBR0UDKjQMgFUU ]

Posts61
Comments147
View on HN
lofi.so 1y ago

Comment with your favorite local-first content

yonz
2pts3
alexop.dev 1y ago

What is Local first development

yonz
119pts55
skyfirehose.com 1y ago

Bluesky Jetstream with DuckDB

yonz
3pts0
ospfranco.com 1y ago

Generating iOS libs by op-SQLite author

yonz
2pts2
news.ycombinator.com 2y ago

Ask HN: Xor Operation for Embeddings

yonz
1pts0
wrily.foad.me.uk 2y ago

FedCM Federated Sign-In Beyond Google and FB

yonz
3pts0
twitter.com 2y ago

Would you do it? Filter to delete emails containing "unsubscribe"

yonz
1pts1
news.ycombinator.com 2y ago

Ask HN: Do you auto delete emails?

yonz
2pts2
vella.ai 2y ago

Vella.ai Feedback

yonz
1pts1
bsky.app 2y ago

Martin Kleppmann talk on local-first (LoFi)

yonz
111pts27
news.ycombinator.com 2y ago

Ask HN: Kubernetes ingress re-write for FEs

yonz
1pts0
datajenius.com 2y ago

NLP Embeddings from 1hot to Word2Vec to Bert

yonz
2pts1
news.ycombinator.com 2y ago

Tell HN: TablePlus is a slick GUI db editor

yonz
3pts0
news.ycombinator.com 2y ago

Ask HN: Extract Any Date Format

yonz
1pts3
twitter.com 2y ago

Memex like products/communities for gathering personal data?

yonz
2pts1
etherscan.io 2y ago

Accidental $20M crypto bounty 0x0000

yonz
3pts2
documents.pub 2y ago

Elegant Palantir Access Control PowerPoint (2008)

yonz
3pts2
news.ycombinator.com 2y ago

Ask HN: How to keep pointer position on mission control? [SOVED]

yonz
1pts1
news.ycombinator.com 2y ago

Ask HN: Scam from `service@paypal.com` email, How?

yonz
7pts7
bugs.chromium.org 3y ago

Android Chrome doesn't support autofill framework

yonz
1pts0
tailscale.com 3y ago

Tail scale NAT traversal (2020)

yonz
1pts1
hackmd.io 3y ago

Explainer blogs for IPFS things (IPSL,DHT, Slices) by Jorropo

yonz
8pts0
news.ycombinator.com 3y ago

Ask HN: Discord broadcast keeps f'n up. Any tips?

yonz
1pts1
twitter.com 3y ago

LoFi (Local First) Meetup #5 at 5PM ET Tomorrow (June 27)

yonz
2pts1
hackaday.com 3y ago

Linux dynamic kernel buffer explot via loop_rw_ITER in fs/io_uring.c

yonz
2pts2
news.ycombinator.com 3y ago

Ask HN: List your paid apps that you could build in a month

yonz
2pts0
developers.facebook.com 3y ago

Replacing Electron with react-native-destop – FB Messenger

yonz
3pts1
github.com 3y ago

Practical Guide for LLMs

yonz
1pts0
discord.com 3y ago

Local First Meetup: James Arthur CEO of ElectricSQL – Tuesday 2PM ET

yonz
2pts1
news.ycombinator.com 3y ago

Ask HN: SSD File Transfer Speeds

yonz
2pts1

It's so fishy, I spent the morning reading sam'AMA and it's a classic whitewashing act. OpenAI is claiming their setup is stronger and that DOW has agreed to their red lines but read the agreement below, it only says use in compliance with laws and executive order.

Anthropic wouldn't have walked away from a multi million contract if their two redlines could be respected. OpenAI on the other hand is a fast, willing and ready company. I would love to see Anthropic's proposed contract

In our agreement, we protect our red lines through a more expansive, multi-layered approach. We retain full discretion over our safety stack, we deploy via cloud, cleared OpenAI personnel are in the loop, and we have strong contractual protections. This is all in addition to the strong existing protections in U.S. law.

We believe strongly in democracy. Given the importance of this technology, we believe that the only good path forward requires deep collaboration between AI efforts and the democratic process. We also believe our technology is going to introduce new risks in the world, and we want the people defending the United States to have the best tools.

Our agreement includes:

1. Deployment architecture. This is a cloud-only deployment, with a safety stack that we run that includes these principles and others. We are not providing the DoW with “guardrails off” or non-safety trained models, nor are we deploying our models on edge devices (where there could be a possibility of usage for autonomous lethal weapons).

Our deployment architecture will enable us to independently verify that these red lines are not crossed, including running and updating classifiers.

2. Our contract. Here is the relevant language:

The Department of War may use the AI System for all lawful purposes, consistent with applicable law, operational requirements, and well-established safety and oversight protocols. The AI System will not be used to independently direct autonomous weapons in any case where law, regulation, or Department policy requires human control, nor will it be used to assume other high-stakes decisions that require approval by a human decisionmaker under the same authorities. Per DoD Directive 3000.09 (dtd 25 January 2023), any use of AI in autonomous and semi-autonomous systems must undergo rigorous verification, validation, and testing to ensure they perform as intended in realistic environments before deployment.

For intelligence activities, any handling of private information will comply with the Fourth Amendment, the National Security Act of 1947 and the Foreign Intelligence and Surveillance Act of 1978, Executive Order 12333, and applicable DoD directives requiring a defined foreign intelligence purpose. The AI System shall not be used for unconstrained monitoring of U.S. persons’ private information as consistent with these authorities. The system shall also not be used for domestic law-enforcement activities except as permitted by the Posse Comitatus Act and other applicable law.

I'll add to this gratitude thread.

Martin has had a material impact on my career. I don't think I would have gotten my job at LinkedIn if it wasn't for his book, Designing Data-Intensive book: https://www.oreilly.com/library/view/designing-data-intensiv...

I learned how to build a web application with FE, BE, DB, and distributed workers for my first job. But it wasn't until I read his book that I understood the enormous gap between building web apps and planet-scale web applications. The book saved me from bombing my interviews.

Well said. After linear.app showed the jump in UX for local data ('moving databases as close to users as possible') it became really easy to answer the questions below by saying checkout linear.app

* Why do we need local-first? * We don't need offline support * Google already has all my data, don't care about data ownership (a little orthogonal)

I know, this has come up a few times. We will have a matrix bridge soon. Do you have any leads for a peer first community software with good adoption?

I have a short list that I'm testing but none of them have the features we need: 1) Live video 2) # channels and 3) adoption

Even moving off Twitter has been a challenge but I am optimistic about bsky

Recently ran into Memex a personal data repository built by Andrew (Brew conf 2018)and realized maybe I'm not alone in wanting to gather structure and explore my personal data from all apps and activities.

Please share if you have hacks, tools, communities or products that can help archive the above.

Tweet thread yielded the following:

- Brew conf 2018 Memex presentation: https://youtu.be/DFWxvQn4cf8?feature=shared

- Chronicle ETL : a cli for ETL'ing from different data siloes https://github.com/chronicle-app/chronicle-etl

- Mainframe https://github.com/andreterron/mainframe

- Dogsheep is a collection of tools for personal analytics using SQLite and Datasette. https://dogsheep.github.io/