Hacker News: veunes

New comment by veunes in "The Chinese whiz kids of Silicon Valley"

veunes — Wed, 13 May 2026 10:17:48 +0000

What stood out to me is how little of this is really new to Silicon Valley

New comment by veunes in "China is going dark to develop its own Mythos"

veunes — Wed, 13 May 2026 10:15:14 +0000

If a model like that exists behind closed doors, we're unlikely to know from press releases

New comment by veunes in "China is going dark to develop its own Mythos"

veunes — Wed, 13 May 2026 10:13:36 +0000

The interesting failure mode here is that the defensive side needs access before the offensive side starts using the capability

New comment by veunes in "The price of power – why batteries are starting to look obvious"

veunes — Wed, 13 May 2026 10:10:25 +0000

The wholesale spread makes batteries look obvious, but I'd be careful about extrapolating from today's arbitrage. If enough batteries get built, the spread they rely on compresses

New comment by veunes in "Sam Altman Testifies That Elon Musk Wanted Control of OpenAI"

veunes — Wed, 13 May 2026 10:08:00 +0000

This sounds less like a dispute about "the mission" and more like a dispute about who should have controlled the mission. The nonprofit/for-profit structure is messy, but so is the idea that putting OpenAI under Tesla would somehow have preserved its independence

New comment by veunes in "Instagram is dropping end-to-end encrypted chats"

veunes — Fri, 08 May 2026 13:38:41 +0000

This is a good reminder that encrypted and end-to-end encrypted are very different promises

New comment by veunes in "Robot Monk at a Buddhist Temple in South Korea"

veunes — Fri, 08 May 2026 13:34:35 +0000

This is a lot less strange if you think of the robot as participating in ritual rather than "having faith"

New comment by veunes in "Movie poster artist Tony Stella dies"

veunes — Fri, 08 May 2026 13:30:07 +0000

45 is painfully young

New comment by veunes in "Belgium stops decommissioning nuclear power plants"

veunes — Thu, 30 Apr 2026 16:20:14 +0000

The manipulation risk is real, but it usually comes from pretending there is a painless answer

New comment by veunes in "Belgium stops decommissioning nuclear power plants"

veunes — Thu, 30 Apr 2026 16:17:05 +0000

The "old car" analogy seems right, with the extra complication that the car is supplying a non-trivial chunk of the country's electricity and replacing it is not quick

New comment by veunes in "Belgium stops decommissioning nuclear power plants"

veunes — Thu, 30 Apr 2026 15:04:38 +0000

The interesting part will be whether Belgium can turn this into a coherent long-term plan

New comment by veunes in "An AI agent deleted our production database. The agent's confession is below"

veunes — Mon, 27 Apr 2026 14:55:22 +0000

I bet if you could look at the hidden reasoning tokens at the exact moment the DB was dropped, there were zero thoughts about safety rules in there. The model simply hit an access error > searched for a token > found one > ran the command. That whole "I am violating my instructions" vector only fired up after the pissed-off user fed it a prompt full of accusations. So yeah, it's not a confession at all, it's just the model adapting to the user's context

New comment by veunes in "An AI agent deleted our production database. The agent's confession is below"

veunes — Mon, 27 Apr 2026 13:44:50 +0000

"backups in the same volume" aren't backups, they’re just snapshots in the same blast radius fwiw. If your DR plan hinges on a single physical volume ID, you have zero resilience

This needs to be a lesson for everyone: real backups belong in an independent store (S3/GCS) in a different region with object lock enabled. It’s the only way to make sure even a compromised root token can’t nuke your data for 30 days

New comment by veunes in "Show HN: Gemma Gem – AI model embedded in a browser – no API keys, no cloud"

veunes — Wed, 15 Apr 2026 16:21:40 +0000

CORS protects your Facebook from your Gmail, but it won't protect your Gmail from the agent itself since it already has access to the DOM and JS context. If that agent gets hit with a prompt injection and decides to "Delete all mail" or exfiltrates session tokens to a third-party endpoint, the browser sandbox will actually facilitate it because it views those as legitimate user-initiated actions

New comment by veunes in "Google Gemma 4 Runs Natively on iPhone with Full Offline AI Inference"

veunes — Wed, 15 Apr 2026 15:58:07 +0000

Sure, 26B models on beefy desktop silicon are finally nipping at the heels of commercial APIs, but this is a mobile thread. On a phone with 8GB of RAM and passive cooling, your tokens per second (t/s) are going to fall off a cliff after the first minute of sustained compute

New comment by veunes in "Google Gemma 4 Runs Natively on iPhone with Full Offline AI Inference"

veunes — Wed, 15 Apr 2026 15:49:32 +0000

It’s likely a llama.cpp backend issue. On the Pixel, inference hits QNN or a well-optimized Vulkan path that distributes the SoC load properly. On the iPhone, everything is shoved through Metal, which maxes out the GPU immediately and causes instant overheating. Until Apple opens up low-level NPU access to third-party models, iPhones will just keep melting on long-context prompts

New comment by veunes in "Google Gemma 4 Runs Natively on iPhone with Full Offline AI Inference"

veunes — Wed, 15 Apr 2026 15:32:41 +0000

This article is all fluff because real benne marketing. If they mentioned that a 4B model on an iPhone 16 drains 15% of the battery for a single long prompt and triggers hard thermal throttling after 20 seconds, nobody would be clicking on headlines about "commercial viability" fwiw

New comment by veunes in "Google Gemma 4 Runs Natively on iPhone with Full Offline AI Inference"

veunes — Wed, 15 Apr 2026 15:27:50 +0000

I noticed the inference is routed through the gpu rather than the Apple neural engine. Google’s engineers likely gave up on trying to compile custom attention kernels for Apple’s proprietary tensor blocks iirc. While Metal is predictable and easy to port to, it drains the battery way faster than a dedicated NPU. Until they rewrite the backend for the ANE, this is just a flashy tech demo rather than a production-ready tool

New comment by veunes in "What being ripped off taught me"

veunes — Mon, 06 Apr 2026 14:55:20 +0000

Yeah, a lot of it only becomes obvious in hindsight because each individual signal is easy to rationalize away

New comment by veunes in "What being ripped off taught me"

veunes — Mon, 06 Apr 2026 14:54:13 +0000

You're not just delivering expertise, you're stepping into a situation where incentives are already misaligned, expectations are fuzzy, and there's often a cashflow problem hiding somewhere