Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

GPT-4o

openai.com

Post reply on HN
Print view View original

541–550 of 1001 posts

Re: GPT-4o

#541

Post by nojvek »

ryankrage77 wrote: post #538
erickhill wrote: post #413

I think it’s safe to say Siri and Alexa are officially dead. They look like dusty storefront mannequins next to Battlestar replicants at this point.

This can't set alarms, timers, play music, etc. The only current overlapping use case I see is checking the weather (assuming GPT-4o can search online), and Siri is already fine for that. Amazing tech, but still lacking in the integrations I'd want to use voice for.

Very easy to plug in that capability with tool use. Gpt3+ already support using tools/json schema output.

Re: GPT-4o

#542

Post by teleforce »

Nobody in the comments seems to notice or care about GPT-4o new additional capability for performing searches based on RAG. As far as I am concerned this is the most important feature that people has been waiting for ChatGPT-4 especially if you are doing research. By just testing on one particular topic that I'm familiar with, using GPT-4 previously and GPT-4o the quality of the resulting responses for the latter is very promising indeed.

Re: GPT-4o

#543

Post by cal85 »

lobochrome wrote: post #511
cal85 wrote: post #495

We've had voice input and voice output with computers for a long time, but it's never felt like spoken conversation. At best it's a series of separate voice notes. It feels more like texting than talking. These demos show people talking to artificial intelligence. This is new. Humans are more partial to talking than writing. When people talk to each other (in person or over low-latency audio) there's a rich metadata…

I don’t know. Have you even seen a gen z?

I don’t follow, what about them?

Re: GPT-4o

#544

Post by px43 »

mvkel wrote: post #238
Jensson wrote: post #24

The most impressive part is that the voice uses the right feelings and tonal language during the presentation. I'm not sure how much of that was that they had tested this over and over, but it is really hard to get that right so if they didn't fake it in some way I'd say that is revolutionary.

I was in the audience at the event. The only parts where it seemed to get snagged was hearing the audience reaction as an interruption. Which honestly makes the demo even better. It showed that hey, this is live. Magic.

I wonder when it will be able to understand that there is more than one human talking to it. It seems like even in today's demo if two people are talking, it can't tell them apart.

Re: GPT-4o

#545

Post by oersted »

teleforce wrote: post #542

Nobody in the comments seems to notice or care about GPT-4o new additional capability for performing searches based on RAG. As far as I am concerned this is the most important feature that people has been waiting for ChatGPT-4 especially if you are doing research. By just testing on one particular topic that I'm familiar with, using GPT-4 previously and GPT-4o the quality of the resulting responses for the latter is…

Can you be more specific? I can’t find this in the announcement. How does this work? What example did you try?

EDIT: web search does seem extremely fast.

Re: GPT-4o

#546

Post by nilsherzig »

Imagine having to interact with this thing in an environment where it is in the power position.

Being in a prison with this voice as your guard seems like a horrible way to lose your sanity. This aggressive friendlyness combined with no real emotions seems like a very easy way to break people.

There are these stories about nazis working at concentration camps, having to drink an insane amount of alcohol to keep themselves going (not trying to excuse their actions). This thing would just do it, while being friendly at the same time. This amount of hopeless someone would experience if they happen to be in custody of a system like this is truly horrific.

Re: GPT-4o

#547

Post by robwwilliams »

laplacesdemon48 wrote: post #411

I recently subscribed to Perplexity Pro and prior to this release, was already strongly considering discontinuing ChatGPT Premium. When I first subscribed to ChatGPT Premium late last year, the natural language understanding superiority was amazing. Now the benchmark advances, low latency voice chat, Sora, etc. are all really cool too. But my work and day-to-day usage really rely on accurately sourced/cited informati…

In the same situation as you. Genomics data mining with validated LMM responses would be a godsend. Even more so when combined with rapid conversational interactions.

We are not far from the models asking themselves questions. Recurrence will be ignition = first draft AGI. Strap in everybody.

Re: GPT-4o

#549

Post by Quarrelsome »

jimkleiber wrote: post #508

I worry that this tech will amplify the cultural values we have of "good" and "bad" emotions way more than the default restrictions that social media platforms put on the emoji reactions (e.g., can't be angry on LinkedIn). I worry that the AI will not express anger, not express sadness, not express frustration, not express uncertainty, and many other emotions that the culture of the fine-tuners might believe are "bad…

Imagine how warped your personality might become if you use this as an entire substitute for human interaction. Should people use this as bf/gf material we might just be further contributing to decreasing the fertility rate.

However we might offset this by reducing the suicide rate somewhat too.

Re: GPT-4o

#550

Post by zedin27 »

I am not fluent in Arabic at all, and being able to use this as a tool to have a conversation will make it more dependent. We are approaching a new era where we will not be "independently" learning a language but ignore the fact of learning it beforehand. Double-edged sword cases
Post reply on HN