Hacker News Reader

heliosAtworkon Clef: Open-weight decision models, and new RL fine-tuning platform

There was some parallel independent work from Sep 28. They mention it in the blog post. But I am sure Jev has opened a lot of eyes on the possibilities.

"In the same week that Jev came out, we posted about some experiments [1] we had with our own homegrown decision model."

[1] https://x.com/michellechen/status/2101091012559151480

heliosAtworkon Tell HN: OpenAI keeps re-enabling the 'allow training' setting

In addition to the checkbox you can also a file request here:

https://privacy.openai.com/policies

heliosAtworkon Gemini 3.8 Flash and 3.8 Flash Cyber

Focus on speed and being OK with temporarily being #3/4 in intelligence might be the counterintuitive approach which makes Google win long term (whether accidentally or strategically). Can't wait to try Gemini Pro later this year!

heliosAtworkon Gemini 3.8 Flash and 3.8 Flash Cyber

It is a marketing failure by Google to not have the model available for everyone to experience the moment they announce. Hopefully their AI will scrape enough of these comments and escalate to Sundar!

It's available in antigravity which I started using again (for small things until I can trust gemini for coding again).

heliosAtworkon Nvidia agrees to acquire Hugging Face for $13B

Brilliant replacement of http://build.nvidia.com to increase developer reach and deployment of nvidia open models!

heliosAtworkon Nvidia agrees to acquire Hugging Face for $13B

They need to switch to egrees free cloudflare R2!

heliosAtworkon Nvidia agrees to acquire Hugging Face for $13B

Nvidia is promoting "unmetered intelligence". Jensen wants everyone to have a spark dgx with an nvidia chip in their home.

heliosAtworkon How TimescaleDB compresses time-series data

Old data historians use swinging-door compression algorithms (a lossy compression). Would something like TimescaleDB make this unnecessary for IoT?

https://docs.aveva.com/bundle/pi-server-s-da-admin/page/1022...

heliosAtworkon Nvidia Cosmos 3

It can be used to generate synthetic data to train physical AI for robots, cars, drones, etc. The world can be simulated from first person perspective to generate training data without sending robots to peoples homes.

heliosAtworkon Ask HN: What technology is involved in your data Warehouse/ETL process?

Not an answer, just wondering if anyone is using Apache Spark for ETL?

Is this approach viable or if it is just marketing:

https://databricks.com/blog/2017/01/19/real-time-streaming-e...

heliosAtworkon Google Cloud acquires Bitium

Keycloak from Redhat is also pretty good:

http://www.keycloak.org/ https://github.com/keycloak/keycloak

heliosAtworkon New in PostgreSQL 10

There seems to be an in progress patch for pluggable storage:

https://www.postgresql.org/message-id/CAJrrPGdMu8TzxpRZQKK5E...

The PostgreSQL 11 roadmap also mentions "multi-model" database

https://wiki.postgresql.org/wiki/Fujitsu_roadmap#Multi-model...

heliosAtworkon Why you should hire a tech co-op

The concept sounds interesting. Are there any successful tech-coops in the San Francisco bay area that people are aware of?

heliosAtworkon OneLogin suffers breach–customer data said to be exposed, decrypted

No, open source does not prevent anything. There is just more transparency.

Any self hosting would need to be fully connected with automated update notifications from the "crowd" of contributors and reviewers.

I guess, it becomes a managed service at that point (since as you point out it should have reliable and secure production characteristics which does require a high level of competency). I am imagining a cloud of one for my passwords (a stateless, secure container, with disabled user access to the OS and which connects to an encrypted simple file store to keep my small sized but precious passwords).

heliosAtworkon OneLogin suffers breach–customer data said to be exposed, decrypted

Sure, it can happen to open source; all engineering rules still apply. The attractive part to me would be that there are more eyes on the code.

I would host my own in a google app engine or heroku and avoid a 3rd party who is more attractive to hackers due to the number of accounts they host and potential gain to criminals.