Live data from Hacker News

Learning to Reason with LLMs

openai.com

731–740 of 1001 posts

Re: Learning to Reason with LLMs

#731

I have a question. The video demos for this all mention that the o1 model is taking it's time to think through the problem before answering. How does this functionally differ from - say - GPT-4 running it's algorithm, waiting five seconds and then revealing the output? That part is not clear to me.

It is recursively "talking" to itself to plan and then refine the answer.

Re: Learning to Reason with LLMs

#732
Maybe this is improvement in some areas, still I got spurious reasoning and inability to connect three simple facts:

Yes, "el presente acta de nacimiento" is correct in Spanish.

Explanation:

    "Acta" is a feminine noun that begins with a stressed "a" sound. In Spanish, when a feminine singular noun starts with a stressed "a" or "ha", the definite article "la" is replaced with "el" to facilitate pronunciation. However, the noun remains feminine.

    Adjectives and modifiers that accompany the noun "acta" should agree in feminine gender and singular number. In this case, "presente" is an adjective that has the same form for both masculine and feminine singular nouns.
So, combining these rules:

    "El" (definite article used before feminine nouns starting with stressed "a")
    "Presente" (adjective agreeing in feminine singular)
    "Acta de nacimiento" (feminine noun with its complement)
Therefore, "el presente acta de nacimiento" is grammatically correct.

Re: Learning to Reason with LLMs

#734
post #518

This is incredible. In April I used the standard GPT-4 model via ChatGPT to help me reverse engineer the binary bluetooth protocol used by my kitchen fan to integrate it into Home Assistant. It was helpful in a rubber duck way, but could not determine the pattern used to transmit the remaining runtime of the fan in a certain mode. Initial prompt here [0] I pasted the same prompt into o1-preview and o1-mini and both c…

What's the incredible part here? Being able to write code to turn hex into decimal?

Also, if you actually read the "chain of thought" contains several embarrassing contradictions and incoherent sentences. If a junior developer wrote this analysis, I'd send them back to reread the fundamentals.

Re: Learning to Reason with LLMs

#735
post #699

This is incredible. In April I used the standard GPT-4 model via ChatGPT to help me reverse engineer the binary bluetooth protocol used by my kitchen fan to integrate it into Home Assistant. It was helpful in a rubber duck way, but could not determine the pattern used to transmit the remaining runtime of the fan in a certain mode. Initial prompt here [0] I pasted the same prompt into o1-preview and o1-mini and both c…

What is the brand of the fan? Same problem here with proprietary hood fan...

InVENTer Pulsar

Re: Learning to Reason with LLMs

#736

For folks who want to see some demo videos and be amazed! HTML Snake - https://vimeo.com/1008703890 Video Game Coding - https://vimeo.com/1008704014 Coding - https://youtu.be/50W4YeQdnSg?si=IohJlJNY-WS394uo Counting - https://vimeo.com/1008703993 Korean Cipher - https://vimeo.com/1008703957 Devin AI founder - https://vimeo.com/1008674191 Quantum Physics - https://vimeo.com/1008662742 Math - https://vimeo.com/10087041…

In "HTML Snake" the video cuts just as the snake intersects with the obstacle. Presumably because the game crashed (I can't see endGame defined anywhere) This video is featured in the main announcement so it's kinda dishonest if you ask me.

Seeing this makes me wonder if they have frontend \ backend engineers working on code, because they are selling the idea that the machine can do all that, pretty hypocritical for them if they do have devs for these roles.

Re: Learning to Reason with LLMs

#737

Student here. Can someone give me one reason why I should continue in software engineering that isn't denial and hopium?

Most of these posts are from romantics. Software engineering will be a profession of the past, similar to how industrial jobs hardly exist. If you have a strong intuition with software & programming you may want to shift towards applying AI into already existing solutions.

The question is, why wouldn't nearly all other white collar jobs be professions of the past as well? Does the average MBA or whatever possess some unique knowledge that you couldn't generate with an LLM fed with company data? What is the alternative career path?

I think software engineers who also understand business may yet have an advantage over pure business people, who don't understand technology. They should be able to tell AI what to do, and evaluate the outcome. Of course "coders" who simply produce code from pre-defined requirements will probably not have a good career.

Re: Learning to Reason with LLMs

#738

This is incredible. In April I used the standard GPT-4 model via ChatGPT to help me reverse engineer the binary bluetooth protocol used by my kitchen fan to integrate it into Home Assistant. It was helpful in a rubber duck way, but could not determine the pattern used to transmit the remaining runtime of the fan in a certain mode. Initial prompt here [0] I pasted the same prompt into o1-preview and o1-mini and both c…

What if you copy the whole reasoning process example provided by OpenAI, use it as a system prompt (to teach how to reason), use that system prompt in Claude, got4o etc?

Re: Learning to Reason with LLMs

#739

This is incredible. In April I used the standard GPT-4 model via ChatGPT to help me reverse engineer the binary bluetooth protocol used by my kitchen fan to integrate it into Home Assistant. It was helpful in a rubber duck way, but could not determine the pattern used to transmit the remaining runtime of the fan in a certain mode. Initial prompt here [0] I pasted the same prompt into o1-preview and o1-mini and both c…

I'm impressed. I had two modified logic puzzles where ChatGPT-4 fails but o1 succeeds. The training data had too many instances of the unmodified puzzle, so 4 wouldn't get it right. o1 manages to not get tripped up by them.

https://chatgpt.com/share/66e35c37-60c4-8009-8cf9-8fe61f57d3...

https://chatgpt.com/share/66e35f0e-6c98-8009-a128-e9ac677480...

Post reply on HN