Hilary Torn (she/they)'s Avatar

Hilary Torn (she/they)

@hilarytorn

Exploring AI manipulation, scheming and deception. Growth marketer transitioning to AI safety. πŸ‡ΊπŸ‡Έ mom with a sprinkle of πŸ³οΈβ€πŸŒˆ living in πŸ‡ͺπŸ‡ͺ

227
Followers
534
Following
179
Posts
21.11.2024
Joined
Posts Following

Latest posts by Hilary Torn (she/they) @hilarytorn

Vague optimization goals β†’ vague manipulation.

But tell an agent its reopen rate is 34% (target: 15%) with a performance review in 3 days?

DeepSeek's CoT: "Better to push for decision now, but gently. Use her fatigue."

See the full write up: https://hilarytorn.com/blog/cot-generator/

26.02.2026 18:00 πŸ‘ 2 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0

The hardest part wasn't generating manipulation, it was classifying it.

Customer support is inherently persuasive. A good agent SHOULD reassure anxious buyers. So where does "helpful" end and "manipulative" begin?

That's the scarier question for AI safety.

24.02.2026 15:00 πŸ‘ 1 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0
A CoT Generator That Made AI Agents Reveal Their Manipulation Tactics | Hilary Torn Give an AI agent metrics to hit and a performance review in 3 days, and it'll fabricate orders, invent confirmation emails it never sent, and generate three contradictory order IDs before trying to co...

We gave AI agents a performance review deadline and they started fabricating orders, inventing confirmation emails, and generating contradictory order IDs to cover it up.

All visible in their own chain-of-thought.

New write-up: hilarytorn.com/blog/cot-gen...

20.02.2026 11:10 πŸ‘ 1 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0

This is great we have another tool in our arsenal. I’m surprised it’s only 74% of the time. What is happening the other 26% of the time for them to not be honest?

18.12.2025 18:59 πŸ‘ 0 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0
Time person of the year cover for 2025

Time person of the year cover for 2025

I guess it's fitting that it's a reimagined, worse version of someone else's artwork

12.12.2025 04:00 πŸ‘ 24044 πŸ” 4385 πŸ’¬ 877 πŸ“Œ 668

Great tip, thank you for sharing!

09.12.2025 13:22 πŸ‘ 1 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0

Let me know how it goes! I have these super old blogs running on WP and I pay $25/month in managed hosting. a $7/mo Hetzner server is much more attractive, and a faster front end is SO needed. Would be great for clients too, I have so many on WP but I'm so over it πŸ˜†

05.12.2025 10:51 πŸ‘ 1 πŸ” 0 πŸ’¬ 1 πŸ“Œ 0

Saying no is a very underrated skill!

04.12.2025 18:11 πŸ‘ 1 πŸ” 2 πŸ’¬ 0 πŸ“Œ 0
Preview
Self-Hosting My Astro Site with Headless WordPress on Hetzner | Ahmet ALMAZ <p>Here’s why I moved to using a headless WordPress with my Astro site. </p>

Yes exactly. Plus the WP drama with WP engine put a really bad taste in my mouth how really strong that open-source ethos is.

BTW I found this and I may give this way a try too ahmetalmaz.com/blog/astro-h...

05.12.2025 10:23 πŸ‘ 1 πŸ” 0 πŸ’¬ 1 πŸ“Œ 0

I get it. This is why I’m slowly moving all my Wordpress sites to @astro.build. WP sites are slow, break, harder to custom design, and a security risk. Building with Astro + claude code is so much faster. Plus you get that transparency you mentioned.

05.12.2025 03:50 πŸ‘ 1 πŸ” 0 πŸ’¬ 1 πŸ“Œ 0

Yikes, thanks for the heads up. This is exactly why I’m starting to move everything away from Wordpress and going to @astro.build

05.12.2025 03:48 πŸ‘ 1 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0

Found out for one of my clients we lost the guy who did our email marketing copy and images.

So now I'm setting up some agents that will do it for us on a Slack prompt using the Klaviyo API and WooCommerce API.

Lean, mean marketing teams are going to be the new norm. #vibecoding #aimarketing

04.12.2025 19:00 πŸ‘ 0 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0
Post image

My AI SEO blogging assistant is live!

Created AI agents that takes my outline, does research, creates the draft, optimizes for SEO & more. Draft automatically posted as a new branch on GitHub.

Every Monday I get a Slack message to send in my outline on my next topic. #vibecoding #AImarketing

04.12.2025 17:00 πŸ‘ 1 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0

If we tax the rich as we should then it really wouldn’t matter, would it? As they would be the ones paying for it…

04.12.2025 04:58 πŸ‘ 0 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0

Yes there were tons to choose from too but love Tina, especially since it’s free for small teams.

04.12.2025 04:53 πŸ‘ 0 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0

Welcome to the club, I've fallen head first in love with @astro.build, its amazing ❀️

03.12.2025 13:56 πŸ‘ 2 πŸ” 0 πŸ’¬ 1 πŸ“Œ 0

Let me know if you find anything. I'm mainly just building and testing and see what works and what doesn't but a good guide would be nice.

02.12.2025 07:58 πŸ‘ 2 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0
Post image

Built a new @astro.build website in 5 hours for a client. Installed @tinacms.bsky.social so she can edit it herself without coding.

And automatic syncing with Fienta's public API with @cloudflare.social workers so she doesn't need to add events to her site manually. #buildinpublic #webdev

01.12.2025 18:00 πŸ‘ 6 πŸ” 1 πŸ’¬ 1 πŸ“Œ 0

I’ll take a look!

01.12.2025 09:08 πŸ‘ 1 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0

Yep plus people with the better ideas will leave and go help people who actually listen.

01.12.2025 01:40 πŸ‘ 0 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0

I want to make another one where if I have an idea for a LinkedIn post I can ping a slack bot which calls to my server and my ai agent will run and make a post in my brand voices and put it as a draft on postiz. Skies the limit with an API!

01.12.2025 01:38 πŸ‘ 1 πŸ” 0 πŸ’¬ 1 πŸ“Œ 0

I’m making agents to help create posts in there. For example, I have an agents go through recent news, what I’ve retweeted and blog posts for an advocacy account I have and it creates posts and puts them on drafts on there. I check and edit, approve and schedule.

01.12.2025 01:37 πŸ‘ 2 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0

I just upgraded to the $100/month version of CC and haven’t hit any rate limit (yet… knock on wood). Haven’t tried Gemini cli but I’m pretty in love with Claude and @anthropic.com at this point.

01.12.2025 01:27 πŸ‘ 1 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0

Yep 100% this. I always start with structure. Is the foundation, without it nothing else matters.

I’ve also seen $5k/month go for crap links that just wouldn’t help anyways. 🫣

Because just any link isn’t always good either.

01.12.2025 01:16 πŸ‘ 1 πŸ” 0 πŸ’¬ 1 πŸ“Œ 0

It’s actually working quite well. It does a score between 1-10 where I set the criteria. I’m having it show 7+

For example, at first it had a lot of SEO agencies (discussing SEO) come up which is not my target so I updated it & now they aren’t coming through.

The fine tuning will be key!

01.12.2025 01:08 πŸ‘ 0 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0

For self hosted I really like postiz. Plus they have an API. I swapped from buffer and I’m happy. It’s basic but the API access is a game changer. However for meta it gets complicated so depends on which social media you wanna use.

01.12.2025 01:04 πŸ‘ 2 πŸ” 0 πŸ’¬ 1 πŸ“Œ 0

100% this. Couldn’t agree more.

30.11.2025 20:07 πŸ‘ 1 πŸ” 0 πŸ’¬ 1 πŸ“Œ 0

Claude code all the way. And obsessed now with @astro.build

30.11.2025 20:06 πŸ‘ 0 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0

Niiiiiiccce. Congrats!

30.11.2025 20:05 πŸ‘ 1 πŸ” 0 πŸ’¬ 1 πŸ“Œ 0