HNHacker News
TopNewBestAskShowJobs

lordofgibbons

2,941 karma · joined January 19, 2020

Hello there!
submissionscomments
lordofgibbons··on Firefox is the best mobile browser
Firefox mobile was unusable slow for most sites I visited and had rendering issues - probably not FF's fault, but webdevs only testing on Chrome. Brave has been very fast with all the spyware and ad blocking features I was looking for in FF. I just had to disable all the crypto stuff first.
lordofgibbons··on Figure 03, our 3rd generation humanoid robot
I have one (a Roborocks), and thinking of getting a second one for the second level of the house. Great value.
lordofgibbons··on Figure 03, our 3rd generation humanoid robot
I know the current early generations of humanoid robots are very janky and slow, but I'd gladly pay $20k USD for a robot to do all the household chores.
lordofgibbons··on Dbos: Durable Workflow Orchestration with Go and PostgreSQL
This is perfect. I've been growing weary of Go projects that can be used as a library, but for some strange reason must to be used as an external service - complicating the infrastructure. Looking forward to using this in a future project.
lordofgibbons··on Self-supervised learning, JEPA, world models, and the future of AI [video]
LeCun has been giving the same talk with literally the exact same slides for the past 3 years. JEPA still hasn't delivered despite FAIR's substantial backing.
lordofgibbons··on GLM-4.6: Advanced Agentic, Reasoning and Coding Capabilies
Very respectable. If Sonnet 4.5 was delayed by a day, GLM 4.6 would have been the obvious best choice for agentic coding... up there with gpt-codex.

The z.ai monthly plan looks like a steal if you're working on personal hobby projects instead of paying the $200/mo price for Claude Code.

lordofgibbons··on Auth.js is now part of Better Auth
I've seen this, and the fact that it's written in Go is kind of irrelevant if I have to manage an external service. I just want it to be simple like what other languages have.
lordofgibbons··on Auth.js is now part of Better Auth
Extremely complex and requires running another multiple services. I just want auth, I don't want to set up kubernetes to orchestrate all the components of Ory.
lordofgibbons··on Auth.js is now part of Better Auth
I really wish there was such an easy off-the shelf auth solution for Go
lordofgibbons··on Game over for pure LLMs. Even Rich Sutton has gotten off the bus
Why do people even listen to Gary Marcus?

His stance on LLMs can be modeled by a simple finite state machine:

State 1) LLM performance stalls for a couple of months: - "See I told you, LLMS are a dead end and won't work!"

State 2) New LLM release makes rapid and impressive improvements - "AI is moving too fast! This is dangerous and we need the government to limit the labs to slow them down!"

Repeat

lordofgibbons··on New Element Web and Desktop apps
What do you recommend as a better (secure/private) alternative?
lordofgibbons··on Microsoft blocks Israel’s use of its tech in mass surveillance of Palestinians
Are you seriously equating observing an area using satellites with indiscriminately monitoring everyone's calls, messages, and possibly hacking their devices?
lordofgibbons··on Unlocking a Million Times More Data for AI
We don't have a data scarcity problem. Further refinement to the pretraining stage will continue to happen, but I don't expect the orders of magnitude of additional scaling to be required any longer. What's lacking is RL datasets and environments.

If any more scaling scaling does happen, it will happen in the mid-training (using agentic/reasoning outputs from previous model versions) and RL training stages.

lordofgibbons··on Cloudflare is sponsoring Ladybird and Omarchy
Maybe a stupid question but: given the massive vulnerability surface area that a browser presents, why would one choose to build it in C++ instead of something memory and concurrency safe like Rust?

I know Rust doesn't automatically make the software safe, but it does rule out a very large % of the exploitable vulnerabilities allowed by unsafe languages like C and C++.

lordofgibbons··on Apple Silicon GPU Support in Mojo
How many people here actually write custom cuda/triton kernels? An extremely small hand full of people write them (and they're all on one discord server).. which then gets integrated into Pytorch, Megatron-LM, vLLM, SGLang, etc. The rest of us in the ML/AI ecosystem have absolutely no incentive to migrate off of python due to network effects even though I think it's a terrible language for maintainable production systems.

If Mojo focuses on systems software ( and gets rid of exceptions - Chris, please <3 ) it will be a serious competitor to Rust and Go. It has all the performance and safety of Rust with a significantly easier learning curve.

lordofgibbons··on Ultrasonic Chef's Knife
That's good point, aerosolized raw chicken meat water doesn't sound like a good time.
lordofgibbons··on As Android developer verification gets ready to go, a new reason to be worried
Yeah, this is exactly my reasoning too. And with Google making it harder for OSes like GrapheneOS to run on Pixel devices, my next device will have to be an Apple one. There's nothing left in the Android ecosystem for me.
lordofgibbons··on DeepSeek writes less secure code for groups China disfavors?
1) Because open-r1 is a toy - a science fair project - not something anyone actually uses. 2) It's based on the techniques described in the R1 paper

The entire open ecosystem in the U.S relies on the generosity of Chinese labs to share their methods in addition to their models.

lordofgibbons··on DeepSeek writes less secure code for groups China disfavors?
Chinese labs are the only game in town for capable open source LLMs (gpt-oss is just not good). There have been talks multiple times by U.S China hawk lawmakers about banning LLMs made by Chinese labs.

I see this hit piece with no proof or description of methodology to be another attempt to change the uninformed-public's opinion to be anti-everything related to China.

Who would benefit the most if Chinese models were banned from the U.S tech ecosystem? I know the public and startup ecosystem would suffer greatly.

lordofgibbons··on Show HN: I reverse engineered macOS to allow custom Lock Screen wallpapers
fair point on the battery life. though my Thinkpad x9 gets a solid 8-10 hours of heavy coding session on a single charge.

> nor does linux wake up from sleep as reliably.

This hasn't been a problem for at least the past 7ish years.

lordofgibbons··on Show HN: I reverse engineered macOS to allow custom Lock Screen wallpapers
If you're going through so much trouble to reverse engineer macOS to customize it in a non-standard way, why even bother staying in the Apple walled prison. Why not just use Linux at that point? You can customize it to your heart's desire.
lordofgibbons··on My first impressions of Gleam
Gleam has caught my eye for the past year or so, and I'd totally learn it if I didn't believe firmly that we won't be coding by hand within the next 9 months. It'll all be done by LLMs so syntax and ergonomics won't mean too much. At least as soon as LLMs learn to stop being turbo-slop generators.
lordofgibbons··on The treasury is expanding the Patriot Act to attack Bitcoin self custody
Bitcoin maximalists are learning that having a non-fungible and fully traceable ledger might be a problem. Even Satoshi called this out! As is, BTC is somewhat of a privacy nightmare. All of your transactions are on the public ledger for anyone with basic knowledge of statistics to correlate and see all of your transactions. Blockchain Analytics is big business!

All the things the Treasury is considering to be "suspicious activity" simply can't be tracked with something that's non-fungible and untracable like Monero. This suspicious activity - aka privacy - is just how all monero transactions are done.

lordofgibbons··on Mistral raises 1.7B€, partners with ASML
> but we have reached the limit of AI scaling with conventional methods

We've just only started RL training LLMs. So far, RL has not used more than 10-20% of the existing pre-training compute budget. There's a lot of scaling left in RL training yet.

lordofgibbons··on Alterego: Thought to Text
Is the video hosted anywhere else? It seems to be the only source of info, but it's being played at like 0.5X speed.
lordofgibbons··on Ex-WhatsApp cybersecurity head says Meta endangered billions of users
Given how WhatsApp is the de-facto way to communicate outside of the West and China, these security/data-handling "weaknesses" are most likely a feature, not a bug. An absolute bonanza for the certain intelligence services.

Remember, kids: End to end encryption is useless if the "ends" are fully controlled by an (untrustworthy) third party.

lordofgibbons··on API Blueprint
Looks to be an abandoned project based on the last Github update being 3 years ago.
lordofgibbons··on ML needs a new programming language – Interview with Chris Lattner
I'm the primary target audiance for Mojo and was very interested in it, but I just wish they didn't keep Exceptions. This backwards compatibility with Python syntax is extremely overrated and not worth the cost of bringing language warts from the 90s.

God, I hate exceptions so much. I have never seen anyone use exceptions correctly in either Java (at FAANG) or in any regular Python application.

I'm much more in favor of explicit error handling like in Go, or the syntax sugar Rust provides.

lordofgibbons··on US to put economic data on 9 blockchains
Even if it's open source but a single entity/foundation has full control of the blockchain, it's a fraud in disguise.
lordofgibbons··on OpenRouter is down
Their uptime is still infinitely better than any single provider though.
← PreviousPage 2 of 14Next →