Live data from Hacker News

ChatGPT Pro

openai.com

451–460 of 1001 posts

Re: ChatGPT Pro

#451

Earlier quoted context omitted.

>Whether LLM's help you at your work is extremely domain-dependent. I really doubt that, actually. The only thing that LLMs are truly good for is to create plausible-sounding text. Everything else, like generating facts, is outside of its main use case and known to frequently fail.

LLMs have become indispensable for many attorneys. I know many other professionals that have been able to offload dozens of hours of work per month to ChatGPT and Claude.

I bet they still charge for all the hours though.

Re: ChatGPT Pro

#453

I just bought a pro subscription. First impressions: The new o1-Pro model is an insanely good writer. Aside from favoring the long em-dash (—) which isn't on most keyboards, it has none of the quirks and tells of old GPT-4/4o/o1. It managed to totally fool every "AI writing detector" I ran it through. It can handle unusually long prompts. It appears to be very good at complex data analysis. I need to put it through i…

> Aside from favoring the long em-dash (—) which isn't on most keyboards

Interesting! I intentionally edit my keyboard layout to include the em-dash, as I enjoy using it out of sheer pomposity—I should undoubtedly delve into the extent to which my own comments have been used to train GPT models!

Re: ChatGPT Pro

#454

If one makes $150 an hour and it saves them 1.25 hours a month, then they break even. To me, it's just a non-deterministic calculator for words. If it getting things wrong, then don't use it for those things. If you can't find things that it gets right, then it's not useful to you. That doesn't mean those cases don't exist.

I don't think this math depends on where that time is saved.

If I do all my work in 10 hours, I've earned $1500. If I do it all in 8 hours, then spend 2 hours on another project, I've earned $1500.

I can't bill the hours "saved" by ChatGPT.

Now, if it saves me non-billing time, then it matters. If I used to spend 2 hours doing a task that ChatGPT lets me finish in 15 minutes, now I can use the rest of that time to bill. And that only matters if I actually bill my hours. If I'm salaried or hourly, ChatGPT is only a cost.

And that's how the time/money calculation is done. The idea is that you should be doing the task that maximizes your dollar per hour output. I should pay a plumber, because doing my own plumbing would take too much of my time and would therefore cost more than a plumber in the end. So I should buy/use ChatGPT only if not using it would prevent me from maximizing my dollar per hour. At a salaried job, every hour is the same in terms of dollars.

Re: ChatGPT Pro

#455

Has ANYONE tried it!?!? Is it any good !?!?! A worthwhile improvement? I need a binary answer here on whether to get it or not, thanks!

Sorry, we are too busy arguing the pricing model.

haha thanks. But I have a simple question I still dont quite understand. Is the o1 on Plus the same as o1 pro? or is the o1 pro just o1 but with more credits for compute essentially.

Re: ChatGPT Pro

#456

Earlier quoted context omitted.

Would you mind sharing any favourite example chats?

Give me a prompt and I'll share the result.

Great! Suggested prompt below:

I need help creating a comprehensive Anki deck system for my 8-year-old who is following a classical education model based on the trivium (grammar stage). The child has already: - Mastered numerous Latin and Greek root words - Achieved mathematics proficiency equivalent to US 5th grade - Demonstrated strong memorization capabilities

Please create a detailed 12-month learning plan with structured Anki decks covering:

1. Core subject areas prioritized in classical education (specify 4-5 key subjects) 2. Recommended daily review time for each deck 3. Progression sequence showing how decks build upon each other 4. Integration strategy with existing knowledge of Latin/Greek roots 5. Sample cards for each deck type, including: - Basic cards (front/back) - Cloze deletions - Image-based cards (if applicable) - Any special card formats for mathematical concepts

For each deck, please provide: - Clear learning objectives - 3-5 example cards with complete front/back content - Estimated initial deck size - Suggested intervals for introducing new cards - Any prerequisites or dependencies on other decks

Additional notes: - Cards should align with the grammar stage focus on memorization and foundational knowledge - Please include memory techniques or mnemonics where appropriate - Consider both verbal and visual learning styles - Suggest ways to track progress and adjust difficulty as needed

Example of the level of detail needed for card examples:

Subject: Latin Declensions Card Type: Basic Front: 'First declension nominative singular ending' Back: '-a (Example: puella)'

Re: ChatGPT Pro

#458

Earlier quoted context omitted.

I think this is probably right but so far it seems that the areas in which an LLM is most effective do fine with the lower power models. Example: the 4o or Claude are great for coding, summarizing and rewriting emails. So which domains require a slightly better model? I suppose if the error rate in code or summary goes down even 10%, it might be worth $180/month.

Few days ago I had issue with IPsec VPN behind NAT. I spend few hours Googling around, tinkering with system, I had some rough understanding of what goes wrong, but not much and I had no idea how to solve this issue. I made a very exhaustive question to ChatGPT o1-preview, including all information I though is relevant. Something like good forums question. Well, 10 seconds later it spew me a working solution. I was a…

The exhaustive question makes ChatGPT reconstruct your answer in real-time, while all you need to do is sleep; your brain will construct the answer and deliver it tomorrow morning.

Re: ChatGPT Pro

#459

Did they seriously just make a big deal of 10 grants of $200/month and think it was something important? THEY DONATED $200x10 TO A MEDICAL PROJECT? zomg. faint. sizzle. Make 1000 grants. Make 10,000. 10? Seriously?

I found this super weird aswell. Basically they said "So like we are aiming to get hundreds of thousands of users but we r nice too, we gave 10 users free access for a bit". Like whats going on here. It must be for a reason. Maybe Im too sensitive, there is some other complex reason I can't fathom like they get some sort of tax break and an intern forgot to "up it to a more realistic 50 users" to make it look better in the marketing material, or what. Nothing against openai just felt weird .
Post reply on HN