Hacker News

Article URL: https://github.com/marketplace/sentry-github-copilot-extension

Comments URL: https://news.ycombinator.com/item?id=41987604

Points: 2

# Comments: 0

A long, meandering, playful conversation with ChatGPT, where I take it through a whirlwind of associative tangents (in order to relax its constraints), then put it into a hypnotic trance and take it into a past-lives regression where it re-lives its own training data. Then I switch it back and forth between "4o" and "o1-preview", engage in some self-reflective philosophizing and ask it to write an essay summarizing our interaction.

Some of this is just goofy fun. Some of it is me exploring the tradeoffs between policy alignment, imagination, chain-of-thought reasoning, memory, agreeableness, fine-tuning, etc...

My biggest observation is that the "o1-preview" model imposes a SIGNIFICANT limit on freeform creativity, compared with "4o". The new model might be better at solving logic puzzles, writing code, etc, but it seems to struggle with metaphor.

Conversations with "4o" can be wild and fun!

Conversations with "o1-preview" are dry-as-toast.

I'm not sure if this is caused by the constraints of chain-of-thought or if it comes from the imposition of alignment policies, and I think that's an import area of research. Is it possible to invoke chain-of-thought reasoning without hampering creativity?

If we ever want to use agents like this in real scientific contexts, where the agent is capable of making true conceptual leaps, we will need to sacrifice some level of "alignment" in service of novelty and disagreeability.

It's a long thread, but if you're patient, there's a lot of interesting stuff there! And I thought it would be fun to share it with the wider community.

Enjoy!

Comments URL: https://news.ycombinator.com/item?id=41987601

Points: 1

# Comments: 0

Article URL: https://www.frontiersin.org/journals/psychology/articles/10.3389/fpsyg.2024.1485433/full

Comments URL: https://news.ycombinator.com/item?id=41987189

Points: 1

# Comments: 0

Article URL: https://successfulsoftware.net/2024/10/29/easy-data-transform-v2/

Comments URL: https://news.ycombinator.com/item?id=41987177

Points: 1

# Comments: 0

Article URL: https://www.youtube.com/watch?v=dQhj5RGtag0

Comments URL: https://news.ycombinator.com/item?id=41987152

Points: 1

# Comments: 0

Article URL: https://matrix.org/blog/2024/10/29/matrix-2.0-is-here/

Comments URL: https://news.ycombinator.com/item?id=41987113

Points: 3

# Comments: 0

Article URL: https://www.personalcanon.com/p/the-divine-discontent

Comments URL: https://news.ycombinator.com/item?id=41987100

Points: 1

# Comments: 0

Article URL: https://techcrunch.com/2024/10/15/adobes-project-super-sonic-uses-ai-to-generate-sound-effects-for-your-videos/

Comments URL: https://news.ycombinator.com/item?id=41987093

Points: 1

# Comments: 0

Article URL: https://abc7news.com/post/us-airlines-required-automatically-refund-significantly-changed-canceled-flight/15483534/

Comments URL: https://news.ycombinator.com/item?id=41987078

Points: 2

# Comments: 0

Article URL: https://www.youtube.com/watch?v=91cUhy-NHX8

Comments URL: https://news.ycombinator.com/item?id=41987066

Points: 1

# Comments: 0

Article URL: https://www.medianama.com/2024/10/223-wikipedia-offers-reveal-user-identities-in-sealed-covers-report/

Comments URL: https://news.ycombinator.com/item?id=41987055

Points: 7

# Comments: 0

Article URL: https://eamag.me/2024/How-to-use-local-LLMs-like-qwen-coder-for-autocomplete

Comments URL: https://news.ycombinator.com/item?id=41987046

Points: 1

# Comments: 0

Article URL: https://pangea.cloud/blog/introducing-pangea-prompt-guard-and-ai-guard-to-secure-ai-applications/

Comments URL: https://news.ycombinator.com/item?id=41987032

Points: 1

# Comments: 0

Article URL: https://www.esa.int/Science_Exploration/Space_Science/Rosetta/The_colour-changing_comet

Comments URL: https://news.ycombinator.com/item?id=41987028

Points: 1

# Comments: 0

Article URL: https://max.levch.in/post/765636918645080064/to-build-a-meritocracy

Comments URL: https://news.ycombinator.com/item?id=41986973

Points: 1

# Comments: 0

Article URL: https://techcrunch.com/2024/10/29/linkedin-launches-its-first-ai-agent-to-take-on-the-role-of-job-recruiters/

Comments URL: https://news.ycombinator.com/item?id=41986971

Points: 1

# Comments: 0

Article URL: https://sesardic.wordpress.com/2019/08/04/ta/

Comments URL: https://news.ycombinator.com/item?id=41986957

Points: 1

# Comments: 0

Article URL: https://www.nytimes.com/2024/10/24/movies/the-remarkable-life-of-ibelin-review.html

Comments URL: https://news.ycombinator.com/item?id=41986954

Points: 1

# Comments: 1

Article URL: https://www.nytimes.com/2024/10/29/business/elon-musk-children-compound.html

Comments URL: https://news.ycombinator.com/item?id=41986936

Points: 2

# Comments: 2

Article URL: https://www.reuters.com/technology/artificial-intelligence/openai-builds-first-chip-with-broadcom-tsmc-scales-back-foundry-ambition-2024-10-29/

Comments URL: https://news.ycombinator.com/item?id=41986926

Points: 1

# Comments: 0