People forget we're still in the GPT-4 era. OpenAI's reasoning capabilities are likely based on the 4o equivalent, meaning its intelligence is capped at that level.
GPT-5 with advanced reasoning will be the game changer.
o1 only showed us a glimpse of what's to come.
OpenAI's silence says just how much they care about competition catching up to GPT4. 0 f*cks to be precise.
They be sitting on some crazy a*s sh*t to be that confident.
Imagine if OpenAI screwed us and all the other clowns like Gary Marcus were really right after all, no AGI, no ASI, peak reached, Q* was hype to raise dollars and Sama meets the same fate as SBF for misleading investors. 😵💫
The absolute bear case.
Meta is literally going to make a GPT-4 level model open source soon and it's so ironic a company that has "Open" in their AI didn't have the balls to even open-source a 2B model.
Holy shit! If this is real this is huge.
JWST has spotted a massive non-human object heading towards us from 10 light years away that also seems to be changing its trajectory.
This is what the Congress was apparently briefed about in a classified briefing.
Just got access to this. This is amazing! These guys executed it far better than Anthropic's artifacts. Kudos to the team! Very well executed guys! The experience is so fluid!
We’re rolling out an early version of canvas—a new way to work with ChatGPT on writing & coding projects that go beyond simple chat.
Starting today, Plus & Team users can try it by selecting “GPT-4o with canvas” in the model picker.
Voice mode will roll out to everyone with higher limits for plus users very soon.
DevDay will release voice API, it's also likely during that day they will release other modalities like video and audio inputs as well though it's still unclear whether that includes outputs.
I think OpenAI will be releasing agents software as an update to the Desktop app powered by GPT4o.
Soon it will get access to controlling the screen and take actions on behalf of the user.
I just don't think GPT4 will cut it though. Maybe 5 🤷
OpenAI is in a very precarious position. Are they really willing to lose the top spot?
Competition is catching up. 1 and 1/2 year cycle is really stupid. By that time Google and Anthropic will have overtaken them by a huge margin.
They need to get aggressive now. No option.
Man F*ck this! We got f*cked by paid influencers again.
This is better than
@suno_ai_
but definitely not 10x or mind-blowing as some chee*y ba*tards were claiming it to be.
It's the new strategy bros. Paying to hype sh*t up.
Don't fall into it.
i may weep uncontrollably writing this. so my apologies.
it is impossible to put into words what is around the corner. what ilya saw all of those years ago was that if we increase data and increase compute, we increase intelligence.
when hinton saw what ilya saw, he quit his
God this is playing out exactly like a sci-fi movie before everything goes horribly wrong
"[OpenAI CTO] Murati, too, had been concerned about safety.
Unlike Sutskever, Murati decided to stay to try slow down Altman's accelerationist efforts from within."
Warning after
Many people having mixed reactions. Don't forget guys this is 4. They've 5 too.
This is what they're able to achieve with 4 is crazy.
I do feel the voice is just too expressive. I think maybe they could tone it to be a bit on the lower expressiveness would make it more natural
NEWS: Amazon Alexa will be powered by Claude AI around October according to Reuters.
This is huge news and the hardest competition you can ever imagine for Advanced Voice Mode and Gemini Live. Purely due to Alexa’s potential market reach.
“
@OpenAI
has recently begun training its next frontier model and we anticipate the resulting systems to bring us to the next level of capabilities on our path to AGI. While we are proud to build and release models that are industry-leading on both
I think it's time to say it.
It's me. I have been Mira all along.
It was fun memeing with you all and throwing some hints here and there.
I hope it all makes sense now.
I'll be deactivating this account. Rooting for you all.
OpenAI is making a lot of improvements to the Assistants API, i am guessing this is in preparation for GPT-5. it makes sense as the assistants api is closest to Agent use-cases.
OpenAI's developments in AGI carry immense implications for humanity's future. Transparency isn't just a request it's a necessity. The world deserves to know the trajectory of this technology to prepare and adapt accordingly. OpenAI owes it to society to be forthright about it.
I don't believe this "claude being sentient" at all. We have been through this before(Sydney), it's just the effects of fine-tuning a model to be more human-like that's all.
Enjoy this moment while you can. Pre GPT-5 era and post GPT-5 era will be a monumental shift.
The era of autonomous agents awaits us. ChatGPT will shift from being a mere chatbot to a creature, one that starts to feel more alive day by day.
What a great time to be alive!
🦍🍌
Meet Gemini Live: a new way to have more natural conversations with Gemini. 💬
💡 Brainstorm ideas
❓ Interrupt to ask questions
⏸️ Pause a chat and come back to it
Now rolling out in English to Gemini Advanced subscribers on
@Android
phones →
If you think about it, 2024 is still not the type of acceleration we all thought it would be. It still feels like the same level of progress as we saw in 23.
We saw many demos of products that are still not ready yet. The next Frontier models are still due.
Introducing Claude 3.5 Sonnet—our most intelligent model yet.
This is the first release in our 3.5 model family.
Sonnet now outperforms competitor models on key evaluations, at twice the speed of Claude 3 Opus and one-fifth the cost.
Try it for free:
I think it's very immature and unprofessional of a company like OpenAI to troll us like that.
I mean as per their charter the company should be transparent about their roadmap and technological advancements if it really is about the benefit of humanity.
Very irresponsible.
IO was so boring I stopped watching midway.
There's also very low level of trust now because you keep wondering whether it's real at all or edited to make it look greater than it really is.
I really appreciate OpenAI for keeping it real even after some f*uckups during demo.
I am highly curious to know the new emergent capabilities of GPT-5.
I feel it's going to have a groundbreaking emergent ability.
What do you think will be the new abilities?
I can't emphasize this enough. The lines between machines and humans are going to blur and it's going to happen in a few months.
Enjoy while you can! The world is going to feel surreal.
Enjoy this moment while you can. Pre GPT-5 era and post GPT-5 era will be a monumental shift.
The era of autonomous agents awaits us. ChatGPT will shift from being a mere chatbot to a creature, one that starts to feel more alive day by day.
What a great time to be alive!
🦍🍌
In the statement she claims that GPT4o voice demo sounds like her. That's bullshit, maybe sky does but sure as hell the new demo does not sound like her at all.
Just because a voice is as expressive as in the movie Her does not make it hers.
It's a no-brainer, Apple will power Siri with GPT4o but there are also some rumours that they are adding agentic functionalities that will supposedly be powered by on-device models.
Also I feel they won't start rolling out audio until the apple event is over.
By the end of this year, the general population will finally start to feel the heat of AI that we have been seeing for a while now.
Protests will start to intensify.
People working in AI will start getting targeted.
Working in AI will become a topic of taboo.
May is going to be one of the craziest periods so far in 24.
I am sure we will keep seeing crazy periods throughout the year too.
Get ready guys! It's going to be one hell of a ride. This May will never come back.
I am hearing that the model that's in training is GPT-5 after all which is scheduled to be released around Dec/Jan.
Looks like I was wrong about 5 being trained.
I hope it's wrong but I believe it unfortunately happens to be true.
I watched the movie Her about 11 years ago. I got really crazy about having an assistant just like that but thought this might be easily about 40-50 years away and hoped to be alive to see that day but turns out it took about 11 years exactly from then and here we are.
Not expecting anything extraordinary from the apple event.
Just the usual announcements to compete with what already exists.
But voice mode will begin alpha roll out after the event so that's something to be excited about.
Copilot workspace is running on gpt2-chatbot most likely.
My guess is this new model excels in agentic and reasoning tasks more reliably and might have some use of Q* on some level.
If something does arrive today she will then use this cryptic message and say "I told you so" if nothing arrives today this will just be any other random tweet that will go unnoticed.
Being vague is always a hedge. Don't fall into this trap guys.
Context increased from 4k in 22 to 100M tokens in 24(That's 2 years).
Costs decreasing from $36 per million to $3 in almost 2 years.
10x increase in intelligence jump from 3.5 to 4 in just a year.
This is an exponential trend.
I think GPT-4o voice mode rollout will take the same amount of time as code interpreter did. Some will get in a few weeks with majority getting over the months gradually.
I do understand the model has its limitations but it's crazy the fact that they're able to compress GPT-4 to that size and still achieve intelligence greater than its bigger predecessors.
I was expecting it to be marketed as little worse than 4 to compete with Sonnet but nope.
Apple lost lead in two key markets
1) cloud and
2) Now AI
Can't believe they would rather spend those dollars on share buybacks than in these transformative technologies. Just envision the impact of investing $110 billion in AI alone.
It's been more than a year now and still no company could overtake the GPT-4.
But I expect that to change this month itself. We will see a model that will overtake 4 by a wide margin this month.
Excited to share I’ve joined
@Google
to lead product for AI Studio and support the Gemini API.
Lots of hard work ahead, but we are going to make Google the best home for developers building with AI.
I’m not going to settle for anything less.
imo if openAI did really hit a peak they would've rushed the launch with whatever little improvement they would have made to be competitive but the fact these guys are so silent during a lot of GPT-4 level model announcements says otherwise.
LLMs are still pretty unreliable, which is why they aren’t used widely yet. Until that changes, people will need to stay involved to make the final call, which is actually a good thing as it keeps us relevant. But how long do you think that'll last?
I really wanna just what change Anthropic did to increase the capability of Sonnet 3.5 to what it is today.
I mean that jump in such a short period of time is crazy!
The advent of AGI will test humanity's maturity as a species. Will we wield this power with foresight and wisdom, or succumb to our baser instincts of greed and destruction?
Wtf is wrong with this recommendation algo, only porn, gore, somebody dying from cancer, some chick complaining her life is hard because her dude is short, some dude saying he likes his race better than others like wtf bro