Feed aggregator

tl;dr - the more agency your agent has, the simpler your use case needs to be

Most if not all successful production use cases today are either human-led or human-in-the-loop. Agent-led is possible but requires simplistic use cases.

---

Human-led:

An obvious example is ChatGPT. One input, one output. The model might suggest a follow-up or use a tool but ultimately, you're the master in command.

---

Human-in-the-loop:

The best example of this is Cursor (and other coding tools). Coding tools can do 99% of the coding for you, use dozens of tools, and are incredibly capable. But ultimately the human still gives the requirements, hits "accept" or "reject' AND gives feedback on each interaction turn.

The last point is important as it's a live recalibration.

This can sometimes not be enough though. An example of this is the rollout of Sonnect 3.7 in Cursor. The feedback loop vs model agency mix was off. Too much agency, not sufficient recalibration from the human. So users switched!

---

Agent-led:

This is where the agent leads the task, end-to-end. The user is just a participant. This is difficult because there's less recalibration so your probability of something going wrong increases on each turn… It's cumulative.

P(all good) = pⁿ

p = agent works correctly n = number of turns / interactions

Ok… I'm going to use my product as an example, not to promote, I'm just very familiar with how it works.

It's a chat agent that runs short customer interviews. My customers can configure it based on what they want to learn (i.e. why a customer churned) and send it to their customers.

It's agent-led because

→ as soon as the respondent opens the link, they're guided from there → at each turn the agent (not the human) is deciding what to do next

That means deciding the right thing to do over 10 to 30 conversation turns (depending on config). I.e. correctly decide:

→ whether to expand the conversation vs dive deeper → reflect on current progress + context → traverse a bunch of objectives and ask questions that draw out insight (per current objective)

Let's apply the above formula. Example:

Let's say:

→ n = 20 (i.e. number of conversation turns) → p = .99 (i.e. how often the agent does the right thing - 99% of the time)

That equals P(all good) = 0.99²⁰ ≈ 0.82

So if I ran 100 such 20‑turn conversations, I'd expect roughly 82 to complete as per instructions and about 18 to stumble at least once.

Let's change p to 95%...

→ n = 20 → p = .95

P(all good) = 0.95²⁰ ≈ 0.358

I.e. if I ran 100 such 20‑turn conversations, I’d expect roughly 36 to finish without a hitch and about 64 to go off‑track at least once.

My p score is high. I had to strip out a bunch of tools and simplify but I got there. And for my use case, a failure is just a slightly irrelevant response so it's manageable.

---

Conclusion:

Getting an agent to do the correct thing 99% is not trivial.

You basically can't have a super complicated workflow. Yes, you can mitigate this by introducing other agents to check the work but this then introduces latency.

There's always a tradeoff!

Know which category you're building in and if you're going for agent-led, narrow your use-case as much as possible.

Comments URL: https://news.ycombinator.com/item?id=43763011

Points: 1

# Comments: 0

Hey!

For the past months I've been building an MVP cursor for email and looking to get my first 10 early users.

The project is still in early development, but would love to hear your input on this project.

current features: categorization, auto draft, using llm to edit/add text, auto task creation from email.

under development: keyboard shortcuts, Cursor-like tab navigation, lots of bug-fixes

I'd love to hear what feedback you have, features you'd like to have or if you'd use/buy the product.

Cheers Doru

Comments URL: https://news.ycombinator.com/item?id=43763002

Points: 1

# Comments: 0

Article URL: https://phys.org/news/2025-04-astronomers-planet-rapidly-disintegrating-comet.html

Comments URL: https://news.ycombinator.com/item?id=43762987

Points: 1

# Comments: 0

Article URL: https://www.pewresearch.org/internet/2025/04/22/teens-social-media-and-mental-health/

Comments URL: https://news.ycombinator.com/item?id=43762976

Points: 1

# Comments: 0

Article URL: https://www.cbsnews.com/news/rfk-jr-autism-study-medical-records/

Comments URL: https://news.ycombinator.com/item?id=43762963

Points: 3

# Comments: 0

Article URL: https://www.token.security/blog/secure-cross-account-access-is-tricky-four-common-dangerous-misconceptions

Comments URL: https://news.ycombinator.com/item?id=43762947

Points: 2

# Comments: 0

Article URL: https://workflow-app.notion.site/Staff-engineer-at-Workflow-1d1a7da236fc804d8177dbed38fec990?pvs=4

Comments URL: https://news.ycombinator.com/item?id=43762945

Points: 1

# Comments: 0

Article URL: https://nim-lang.org/blog/2025/04/22/nim-224-2016.html

Comments URL: https://news.ycombinator.com/item?id=43762925

Points: 2

# Comments: 0

Article URL: https://www.rightclicksave.com/article/whatever-happened-to-nfts

Comments URL: https://news.ycombinator.com/item?id=43762924

Points: 1

# Comments: 0

Article URL: https://www.scientificamerican.com/article/nasas-next-major-space-telescope-is-ready-to-launch-trump-wants-to-kill-it/

Comments URL: https://news.ycombinator.com/item?id=43762910

Points: 4

# Comments: 0

Article URL: https://blog.jpillora.com/p/an-ode-to-mastery

Comments URL: https://news.ycombinator.com/item?id=43762905

Points: 1

# Comments: 0

Article URL: https://www.youtube.com/watch?v=Ed_E2crglcw

Comments URL: https://news.ycombinator.com/item?id=43762903

Points: 2

# Comments: 0

Article URL: http://www.redwedgemagazine.com/online-issue/stafford-beer-eudemony

Comments URL: https://news.ycombinator.com/item?id=43762886

Points: 1

# Comments: 0

Article URL: https://www.seangoedecke.com/lying-to-llms/

Comments URL: https://news.ycombinator.com/item?id=43762884

Points: 1

# Comments: 0

Article URL: https://github.com/nwiizo/tfmcp

Comments URL: https://news.ycombinator.com/item?id=43762877

Points: 2

# Comments: 0

Article URL: https://www.theverge.com/cs/features/652590/vietnam-war-viet-cong-us-military-psy-op-comic

Comments URL: https://news.ycombinator.com/item?id=43762859

Points: 1

# Comments: 0

Article URL: https://ashley.dev/posts/fear-of-being-seen/

Comments URL: https://news.ycombinator.com/item?id=43762846

Points: 1

# Comments: 0

The city of Abilene, Texas, is scrambling to restore systems that have been taken offline in response to a cyberattack.

The post Cyberattack Knocks Texas City’s Systems Offline appeared first on SecurityWeek.

Juneau's internet options may be limited with higher prices and slower speeds than most US cities, but residents still have access to a few reliable ISPs.

It's not often we see a VPN developed as more than just a way to hide your IP address and give you some online protection against tracking. So how does the open-source, Mixnet-based NymVPN project stack up?