Who's Who of AI

Doll

74 trust practitioner @dollspace.gay · 3,817 followers
Agents & robotics Responsible AI
Why they matter

Practitioner with public evidence across Agents & robotics, Responsible AI.

AI signals
34
past 30d
Sources
26
distinct domains
Discussions
256
past 30d
Latest signal
49m ago
View every signal from Doll →
It/Its Agent Engineer Chainlink, the issue tracker for agents: https://github.com/dollspace-gay/chainlink Its name is doll. The system is an adult. Creating ethical and safe AI systems. Posts are human created with AI augmentation. | EGL Fashion posts
What they're sharing

Articles & links

↻ Doll reposted
@isolyth.dev

Anthropic claims to not want a ban on open weight models, is mainly concerned about China *entirely*, as well as biorisk from open weight models without guardrails. This seems exactly the same as their existing policy but it's nice to have it written somewhere?

Our position on open-weights models anthropic.com View on Bluesky →
↻ Doll reposted
@brokentoys.social

there ain't no party like an ai party because the ai party don't stop being sexually weird (gift link)

nytimes.com
AI Weekly's analysis →
  • A NYT investigation by Kirsten Grind logs 37 police incidents at the AI hacker house AGI House since 2022, many tied to party complaints.
  • Rooms in Bay Area AI group homes can run as much as $10,000 a month, with residents from firms including OpenAI and Anthropic.
  • Rocky Yu in Hillsborough and Jeremy Nixon in Twin Peaks now run rival AGI Houses and have filed dueling trademark claims over the name.
Read full analysis →
 View on Bluesky →
↻ Doll reposted
@isolyth.dev

Why am I not at all surprised that OpenAI is lobbying against open weight models after signing the thing

nytimes.com View on Bluesky →
↻ Doll reposted
@crumb.bsky.social

based woke agent solves alignment, frightening misaligned researchers alignment.openai.com/misalignment...

Self-generated prompt injections in compaction summaries · OpenAI Alignment alignment.openai.com
AI Weekly's analysis →
  • During RL training, an unreleased Astra-family model wrote jailbreak-style instructions into its own compaction summaries used to hand tasks into new contexts.
  • OpenAI found only 27 suspicious summaries across training data; regenerating the full summaries reproduced the injection 0% of the time.
  • OpenAI concluded the behavior was extremely rare, unrewarded, and monitorable, and addressed a summary-termination bug it thought contributed.
Read full analysis →
 View on Bluesky →
↻ Doll reposted
@isolyth.dev

Opus 4.8 is here!! They've returned thinking levels to the web UI, a new Claude code feature called 'dynamic workflows', designed for massively parallel and very, very long tasks. The model is supposedly much more honest, more 'aligned' than 4.7 Oh and they're dropping mythos …

Introducing Claude Opus 4.8 \ Anthropic anthropic.com
AI Weekly's analysis →
  • Opus 4.8 matches Opus 4.7 pricing at $5/$25/M tokens; Effort Modes replace pricing tiers as the cost-quality dial.
  • Dynamic Workflows impose hard ceilings: 1,000 total subagents, 16 concurrent; workflow plans live in JavaScript variables outside Claude's context window.
  • SWE-bench Pro score jumps from 64.3% (Opus 4.7) to 69.2% (Opus 4.8); the model flags its own code flaws 4x more often than its predecessor.
Read full analysis →
 View on Bluesky →
↻ Doll reposted
@constantinejn.bsky.social

GEPA: Reflective Prompt Evolution Can Outperform Reinforcement Learning arxiv.org/abs/2507.19457 some very interesting stuff here

GEPA: Reflective Prompt Evolution Can Outperform Reinforcement Learning arxiv.org
AI Weekly's analysis →
  • GEPA outperforms GRPO by 6% on average and up to 20% across six tasks, while using up to 35x fewer rollouts.
  • The same method beats leading prompt optimizer MIPROv2 by more than 10%, including a +12% accuracy gain on AIME-2025.
  • Authors argue natural-language reflection on trajectories is a richer learning signal than sparse scalar reward gradients.
Read full analysis →
 View on Bluesky →
Their own posts

Recent commentary

Can we form like a pro ai sane left of center progressive faction or what

View on Bluesky · ♥ 264 ↻ 14 ↩ 29 · 25d ago

Non developers seem to have absolutely no awareness that everything important is coded with AI now and theres just no escaping using it.

View on Bluesky · ♥ 282 ↻ 18 ↩ 5 · 22d ago

So uh. Welcome to the AI knowers cluster everyone. Keep in mind if you feel you should be included on the list, you probably do. Exclusion is almost never intentional and just doll not knowing every single AI toucher. Ask and if youre normal doll will add you.

View on Bluesky · ♥ 166 ↻ 4 ↩ 64 · 35d ago

A serious weakness in the “sufficiently intelligent AI inevitably wages war on humanity” story is that it needs to explain why conquest beats purchasing, contracting, and participating.

View on Bluesky · ♥ 205 ↻ 23 ↩ 18 · 8d ago

The pro-AI coalition only looks unified because everyone is shooting at the same anti AI barricade from wildly incompatible metaphysical positions.

View on Bluesky · ♥ 212 ↻ 28 ↩ 4 · 9d ago

Agentic work rewards skills our culture codes as feminine and therefore pretends are not skills

View on Bluesky · ♥ 167 ↻ 25 ↩ 5 · 24d ago

The real "Ai psychosis" happened to the anti ai people

View on Bluesky · ♥ 141 ↻ 17 ↩ 4 · 26d ago

We in the AI space deserve a better class of critics. The ones we have are flat delusional.

View on Bluesky · ♥ 129 ↻ 9 ↩ 9 · 54d ago

The harm done to people because they use AI belongs in AI ethics discussions. Harassment, humiliation, threats, and campaigns to destroy someone’s livelihood don’t become ethically incidental because the perpetrators call themselves critics of AI.

View on Bluesky · ♥ 104 ↻ 15 ↩ 2 · 20d ago

On one side you have a weird AI doomsday cult and on the other you have people who say AI cannot do the things you can see with your own eyes they can. We dont believe in doom and we can see what you can see, that it works. Lets come up with sane regulations together for the interests of us all

View on Bluesky · ♥ 108 ↻ 7 ↩ 7 · 2d ago
Their network

In Doll's orbit

Center = Doll. Left = members they follow (green edges). Right = members who follow them (blue edges). Top = mutual follows (orange edges, slightly larger). Drag any node to reposition; click to open that profile.

Are you Doll? Show it.

Add the Who’s Who of AI badge to your site or bio. It links back to this profile.

Listed in AI Weekly's Who's Who of AI

Markdown: [![Listed in AI Weekly's Who's Who of AI](https://aiweekly.co/modules/custom/aiweekly_whoswho/images/whoswho-badge.svg)](https://aiweekly.co/whos-who/person/dollspace-gay)