Was GPT-5 Underwhelming? OpenAI Co-founder Leaves, Figure02 Arrives, Character.AI Gutted, GPT-5 2025

Summary

The transcript discusses the current state of AI development, highlighting delays in GPT-5's release and leadership changes at OpenAI, including Greg Brockman's leave and John Schulman's departure to Anthropic. The narrative mentions other AI advancements, like Meta's Llama models and Google's Gemini 1.5 Pro, suggesting OpenAI may be falling behind. It also covers upcoming events like OpenAI's Dev Day and the ongoing lawsuit from Elon Musk against Sam Altman. Additionally, the transcript touches on humanoid robots using OpenAI's technology and the broader AI landscape's shifting dynamics.

Takeaways

• 🚀 The GPT 5 release date is uncertain, with Open AI leadership experiencing changes and departures.
• 🧘 Greg Brockman, co-founder of Open AI, is taking a leave of absence, highlighting the ongoing challenges in achieving safe AGI.
• 🏢 There is a trend of departures among Open AI co-founders, including John Scharff who left for Anthropic, signaling potential alignment issues.
• 📉 Open AI may be lagging in raw intelligence compared to models like Llama with 45 billion parameters, suggesting a need for significant improvement in their next model.
• 🔮 The anticipation for GPT 5 is high, but its release might be delayed beyond the expected timeframe, possibly due to strategic planning around developer events.
• 🤖 The development of agile, autonomous humanoid robots like Fig2 indicates a potential shift towards practical applications of AI beyond just model development.
• 📉 Elon Musk's lawsuit against Sam Altman and Open AI raises questions about the original intentions and current direction of Open AI.
• 🌐 The debate over who will control the future of AI is intensifying, with concerns about the concentration of power in a few large companies.
• 📊 The LM CIS chatbot Arena leaderboard shows a competitive landscape with models like Gemini 1.5 Pro and Claude 3.5 Sonic outperforming others.
• 🔑 There is a growing trend of smaller AGI labs being overshadowed by larger entities with more resources, such as Google acquiring talent from Character AI.
• 🔄 The importance of a data flywheel effect is emphasized, where more data leads to better AI, which in turn can generate more data.

Q & A

What is the significance of the departure of John Scharff from OpenAI to Anthropic?

John Scharff's departure to Anthropic is significant because it indicates a trend of departures among co-founders at OpenAI, and it suggests that he may not have been working with people who were deeply engaged with the alignment work he was focused on, which is an attempt to align machine values with human values.

What does the term 'alignment' refer to in the context of AI?

In the context of AI, 'alignment' refers to the effort to align machine values with human values, ensuring that AI systems behave in ways that are beneficial and safe for humans.

Why is the release of GPT-5 potentially delayed beyond expectations?

The release of GPT-5 could be delayed because OpenAI has announced events focusing on advancements in APIs and dev tools, which implies that they are not planning to release a new model immediately after these events, suggesting a release date more likely at the end of the year or later.

What is the advanced voice mode mentioned in the script, and how is it related to OpenAI's models?

The advanced voice mode is a feature that makes the voice output of AI models incredibly lifelike. It is related to OpenAI's models as it could potentially be a part of their future offerings, enhancing user interaction with their AI systems.

What is the controversy surrounding Elon Musk's lawsuit against Sam Altman and OpenAI?

Elon Musk is suing Sam Altman and OpenAI, accusing them of lying and deceit, claiming that Altman was motivated by greed and that Musk wanted OpenAI to be more open to compete with Google. The lawsuit describes the original OpenAI as a 'spurious venture' and uses strong language to emphasize the accusations.

How does the script suggest that smaller AI labs like Character AI are struggling to compete with larger labs?

The script suggests that smaller AI labs like Character AI are struggling because they are being overshadowed by the scale of data and compute resources that larger labs like Meta and Google have, which is making it difficult for them to develop models that are as intelligent as those of the larger labs.

What is the significance of the LM Cis Chatbot Arena leaderboard for AI models?

The LM Cis Chatbot Arena leaderboard is significant as it provides a ranking of AI models based on their performance in various tasks, offering a comparative measure of their capabilities and helping to identify trends and strengths among different models.

Why might the new Gemini 1.5 Pro version be considered a more justified leader on the LM Cis Chatbot Arena leaderboard compared to GPT-40 Mini?

The new Gemini 1.5 Pro version might be considered a more justified leader because it performs slightly worse than 3.5 Sonic but far better than other models, suggesting a more balanced and comprehensive performance across various tasks.

What is the potential impact of the advanced voice mode on the adoption of AI models by the public?

The advanced voice mode could significantly increase the adoption of AI models by the public because it makes interactions with AI systems more natural and lifelike, potentially leading to hundreds of millions of people using it once it becomes available.

What does the script suggest about the future of AI control and competition among different companies?

The script suggests that the future of AI control is becoming increasingly concentrated among a few large companies, with OpenAI, Meta, and Google leading the way. It raises questions about who will control the future of AI and the implications of these companies' dominance in the field.

Outlines

00:00

🔍 Open AI Leadership and Departures

The script discusses the shifting landscape of AI leadership, with Open AI facing internal changes as Greg Brockman takes a leave of absence and several key figures depart, including John Schan, the former head of alignment. The departures suggest a lack of alignment with the company's direction, especially concerning the critical issue of AI alignment with human values. There is speculation about the capabilities of Open AI's next frontier model, which may have already been trained, and the implications of these departures on the company's progress towards AGI. The script also touches on the lawsuit by Elon Musk against Sam Altman and Open AI, accusing them of deceit and greed, and the broader debate on who will control the future of AI.

05:03

📊 AI Competition and Model Advancements

This paragraph delves into the competitive AI landscape, highlighting the strides made by companies like Meta with their Llama model and the challenges faced by Open AI to keep pace. It discusses the potential release of advanced models like Claude 3.5 and the significance of open-source competition encouraged by the White House. The script also points out the struggles of smaller AI labs like Character AI, which have been overshadowed by tech giants like Google. The discussion includes the performance of various AI models on benchmarks and leaderboards, emphasizing the difficulty in measuring AI intelligence and the nuances of model capabilities.

10:03

🤖 Integration of AI with Robotics and Future Prospects

The final paragraph focuses on the integration of AI with robotics, particularly the adoption of Open AI's models by Figure O2 humanoid robots. It describes the capabilities of these robots, including advanced voice interaction and physical strength equivalent to humans. The script speculates on the potential for a data flywheel effect as these robots perform tasks autonomously and self-correct. It also mentions the use of Weights & Biases for tracking machine learning experiments and the importance of iterating on LLM applications. The paragraph concludes by reflecting on the shifting dynamics in the AI field and expressing gratitude to the audience for their engagement.

Keywords

💡AGI (Artificial General Intelligence)

AGI refers to a theoretical form of artificial intelligence that possesses the ability to understand, learn, and apply knowledge across a wide range of tasks at a level equal to or beyond that of a human. In the video's context, AGI is the ultimate goal that various AI labs are striving towards, with discussions about whether companies like OpenAI are making progress towards achieving it.

💡OpenAI

OpenAI is a research laboratory that focuses on creating and developing friendly artificial general intelligence (AGI). In the script, OpenAI is highlighted as a key player in the AI field, with discussions around its leadership changes, model developments, and the departure of co-founders.

💡Alignment

Alignment in the context of AI refers to the effort to ensure that the goals and values of an AI system are consistent with human values and interests. The script mentions several departures from OpenAI, particularly those from the alignment team, suggesting challenges in this area.

💡GPT (Generative Pre-trained Transformer)

GPT is a type of AI language model developed by OpenAI that is capable of generating human-like text based on the input it receives. The script discusses the evolution of GPT models, with a focus on the anticipated release and capabilities of GPT 5.

💡Co-founder Departures

The script notes a trend of co-founders leaving OpenAI, which may signal internal disagreements or a shift in direction. These departures are significant as they could impact the company's focus and progress towards AGI.

💡Advanced Voice Mode

Advanced Voice Mode is a feature mentioned in the script that is expected to make AI-generated speech more lifelike and interactive. It is an example of how AI is becoming more integrated into user experiences, potentially leading to widespread adoption.

💡LM CIS Chatbot Arena

LM CIS Chatbot Arena is a leaderboard that ranks AI chatbots based on their performance. The script uses this leaderboard to discuss the relative capabilities of different AI models, such as Gemini 1.5 Pro and Claude 3.5 Sonic.

💡Weights and Biases

Weights and Biases is a tool used for tracking machine learning experiments. In the script, it is mentioned as a sponsor and a resource for AI labs, including OpenAI, to improve their models and applications.

💡Data Flywheel

A data flywheel refers to a self-reinforcing cycle where the accumulation of data leads to improvements in AI models, which in turn generate more data. The script suggests that this concept is relevant to the development of AI, particularly in the context of robots like Fig2.

💡Fig2

Fig2 refers to a humanoid robot mentioned in the script that utilizes AI for tasks and interactions. The robot's capabilities, such as speech and physical tasks, are discussed as indicators of progress towards more advanced AGI.

💡Elon Musk

Elon Musk is mentioned in the script in relation to a lawsuit against Sam Altman and OpenAI, alleging deceit and a lack of openness. This highlights the contentious nature of the AI industry and the differing visions of its future.

Highlights

GPT 5 release date is uncertain, with Open AI leadership facing challenges.

Greg Brockman, co-founder of Open AI, is taking a leave of absence until the end of the year.

John Scharf, another co-founder, has left Open AI for Anthropic, indicating a lack of alignment with the company's direction.

There is a trend of departures among co-founders at Open AI, raising concerns about the company's future.

Open AI has begun training its next Frontier Model, but details on its capabilities remain unknown.

Open AI's Dev Day in October suggests that GPT 5 will not be released before the end of November 2024.

Elon Musk is suing Sam Altman and Open AI, alleging deceit and a lack of openness in the company's operations.

Meta and Google are leading in AI development, with smaller labs struggling to compete.

Character AI, a smaller AGI lab, has been overshadowed by Google's acquisition of key talent and IP.

The debate on AI intelligence suggests that scale in data and compute is more important than obscure tricks or arcane knowledge.

LM CIS chatbot Arena leaderboard shows new models like Gemini 1.5 Pro taking the lead in performance.

Open AI's advanced voice mode is incredibly lifelike, suggesting potential for widespread adoption.

Figure O2 humanoid robots are using Open AI models, indicating a practical application of AI technology.

AI Labs are increasingly relying on tools like Weights & Biases to track machine learning experiments.

The shift in AI development suggests that control over the future of AI is becoming more contested.

The release of raw data by LMIS provides insights into model performance and their refusal to answer certain questions.

The new Gemini 1.5 Pro is being tested for its reasoning capabilities, with results expected to be published online.

AI development is moving towards more practical applications, such as autonomous humanoid robots, rather than just theoretical advancements.

Transcripts

as the GPT 5 release date sails further

into the Horizon open AI leadership

splinters meanwhile other AI Labs ship

incrementally smarter models and smaller

AGI efforts like character AI are

swallowed by the Google whale leaving in

my eyes just four companies remaining in

contention for having the most capable

models out there but if we get agile

autonomous humanoid robots like fig2

that could soon count to 50 in one

breath with chat gbt advanced voice

maybe many will take that as a win for

2024 but let's start with the leadership

of open Ai and Greg Brockman is going on

a leave of absence through to the end of

the year calling it his first time to

relax since co-founding the company he

goes on the mission is far from complete

though we still have a safe AGI to build

now of course any one person taking time

away might well be for person reasons

but we have the full-on departure to

anthropic of another one of the

co-founders John schan the reason that

he as the former head of alignment at

openai jump ship to anthropic I think is

given away in this sentence he said he

wants to do research alongside people

deeply engaged with the topics he's most

interested in now it is pretty hard to

read that as saying anything other than

he wasn't working with people who were

deeply engaged with the kind of

alignment work that he was working on

Just for those who don't know alignment

is this attempt to align machine values

with human values but it follows the

departure of the previous head of

alignment Yan Leica and before that the

previous co-head of alignment and

co-founder Ilia satova I'm definitely

starting to notice a trend of Departures

among co-founders at openai so I've

listed chat PT to count the former heads

of alignment at open AI 1 2 3 4 5 6 7 8

9 10 11 12 13 14 15 16 17 18 192 21 22

23 24 25 26 27 28 29 30 31 32 33 34

obviously I was somewhat joking there I

think there's been less than 10 former

heads of alignment but that advanced

voice mode does seem cool I mean even if

open AI models aren't actually getting

smarter and arguably with GPC 40 are

getting slightly Dumber that advanced

voice mode is incredibly lifelike and I

could see hundreds of millions of people

using it at least when it comes out

which seems to be on the Never Never and

here's some even more important context

open aai back in May said that they had

recently begun training its next

Frontier Model now here we are on August

the 6th it would almost certainly have

finished training by now so those key

players at openai would have a rough

sense for its capabilities all of these

departures after they've trained their

latest Frontier Model seems strange and

then we got this about open ai's

so-called Dev day which is starting in

October October and running through to

November I am sure it will be

fascinating but they released this

particular nugget while we know

developers are waiting for our next big

model which we shared has begun training

earlier this year in May these events

will focus on advancements in the API

and are Dev tools taking that at face

value would imply that quote GPT 5 will

not come before November 21st more

likely it means that gbt 5 wouldn't even

come before the end of the year because

why would you release a model just after

you've invited a load of devs to play

about with your tools now yes samman has

recently claimed in the Washington Post

that more advances will soon follow and

will usher in a decisive period in the

story of human society but in recent

months it would be hard to say that open

AI have produced much in the way of

decisive progress and of course all of

that comes as Elon Musk is suing yet

again Sam mman and open AI for what he

says is lying and pery the lawsuit calls

the original open AI a spurious Venture

the language used throughout this

86-page document is hardly subtle musk

claims that samman is doing a long con

his pery and deceit are of Shakespearean

proportions now it does go on and on but

the basic accusation is that Sam ultman

was motivated by greed and Elon Musk

just wanted to have something more open

to compete versus Google so I think musk

and Others May raise an eyebrow when

later in the article samman said making

sure open-source models are readily

available to developers in other nations

will further bolster our advantage

talking of the US and the question that

the article fundamentally raises is who

will control the future of AI well as of

today it looks less and less likely to

be Sam Morman open AI are certainly good

at productizing Ai and the advanced

voice mode as we saw is great search GPT

could make them some money and S is

coming out at some point presumably this

year and at least at the moment the

figuro 2 robot is using an open AI video

language model we'll get back to that in

just a moment but in terms of raw

intelligence open AI feel like they're

falling behind the Llama 3 45 billion

parameter model is already smarter than

GPT 40 and Zuckerberg has recently

committed to 10 times more computing

power to train llama 4 or to put it

another way the next open AI model would

have to be significantly better than GPT

40 just to catch up to the current

state-ofthe-art let alone the

state-ofthe-art when llama 4 comes out

and of course in the meantime even just

this year we might be getting Claude 3.5

Opus from anthropic or even Claude 4

simply put a year is a long time in Ai

and the debate has moved on even the

White House are now encouraging

open-source competition to the light of

open AI Sam Alman meanwhile is still

warning about people stealing key

intellectual property such as model

weights now do forgive me for pointing

out that one of the modules on my

corsera course is about that difference

between open source and open weights

super grateful of course for those 10

reviewers who have kindly left reviews

for this course but don't get me wrong

it's not like meta is having it all its

own way do you remember those Tom Brady

Paris Hilton chatbots that all the

youngsters were apparently going to be

using I think each celebrity was paid

something like $5 million for a few

hours of recordings well apparently

they're now being scrapped and none of

those AI chat Bots amassed a

particularly big following but nor are

things going particularly well for these

smaller AGI Labs like character AI their

product was or is an array of chat Bots

but they were also aiming at AGI and

training Their Own Foundation models

obviously the leaders of character AI

must have been somewhat disappointed by

those New Foundation models because

essentially they've been bought out or

hollowed out by Google not actually

buying a rival company but taking its

key talent and IP but at this point you

might be starting to notice somewhat of

a trend if the incrementally greater

intelligence of new models were down to

obscure tricks or Arcane knowledge then

you'd expect smaller Labs like character

AI to be doing as well as the biggest

Labs but if it was all about sheer scale

of data and compute you'd expect the

leaders to be increasingly well meta and

Google and that is more or less what

we're seeing though of course measuring

that intelligence is quite hard we do

have the LM CIS chatbot Arena

leaderboard in which the new version of

Gemini 1.5 Pro takes the lead at almost

1300 ELO but if you'll notice we have

GPT 40 mini coming third ahead indeed of

Claude 3.5 Sonic which in my own

Benchmark is far and a way ahead if we

just relied on these ELO rankings you'd

think well if open AI can come up with a

tiny model doing almost as good as the

rest they must be doing amazingly but

lmis recently did something great which

was release a batch of raw data showing

comparisons between the models and which

one won and I looked through the dozens

of examples and one Trend emerged claw

3.5 Sonic essentially refused more

requests than GPT 40 Mini even when both

models couldn't perform a task like

creating an image natively GPT 40 mini

gave it I guess more of a go it at least

describe the image that it would create

or in this example when the models were

asked a political question GPC 40 mini

gave a response whereas Claude 3.5 Sonic

just apologized and said it wouldn't

provide analysis now I have noticed that

myself that Claude 3.5 Sonic is more

sensitive than any other model but

that's not a sign of lacking

intelligence so to the the extent that

we're going to call language modeling

intelligence Claude 3.5 Sonic is far

more capable than GPC 40 mini but this

Rance to answer certain questions could

explain the leaderboard rankings now yes

to anyone following the channel I have

been testing the new version of Gemini

1.5 Pro on my simple bench the final

scores will be presented on a website

that I'm hoping to release before the

next video but in the early testing it

performs slightly worse than 3.5 Sonic

but far better than other models so in

that sense this lead B position could be

far more Justified at least than GPT 40

mini for those who haven't heard of my

new reasoning Benchmark humans score

over 90% quite easily whereas models

like the new Gemini 1.5 pro version

score around 25% it would however be

somewhat hypy to say that this is

another step toward AGI as one of the

co-founders of Google deepmind recently

said but whether you think that new

Gemini 1.5 Pro is the best or Claude 3.5

Sonic certainly more and more people are

now shifting their API spend away from

openai openai letting other labs have

the lead for a couple of weeks could be

a timing issue but a couple of months

just makes it seem like they don't have

a reply but as I mentioned at the start

at least open a eyes models are the ones

being chosen by figure O2 these humanoid

robots have an onboard mic and speaker

so hopefully you could chat to them like

you would the new openai advanced voice

mode in other words seamlessly with very

low latency Adcock the founder of figure

said that the default user interface to

our robot will be speech apparently the

robot can work for around 20 hours

straight which is even more than me

reading the latest AI papers its hands

have 16 degrees of freedom and

apparently human equivalent strength so

honestly even though it can speak back

to you you might not want to speak back

to it now apparently these fig2 robots

can perform certain tasks autonomously

and self-correct and that data flywheel

will be in effect and one might well say

that the argument that ubiquitous robot

assistance will arrive before artificial

general intelligence looks more

plausible than ever and speaking of a

data flywheel you may already know that

AI Labs including open AI have used

weights and biases this video sponsor to

track Frontier machine learning

experiments but what you might not know

is that weit and bu now have a weave a

lightweight tour kit to confidently

iterate on llm applications they also

produce free prompt and llm agent

courses on their website and if you

didn't know that you can let them know

that you came from me by using my

customized link and the link is in the

description so in short the vibe is

Shifting but what isn't changing is my

gratitude for you watching all the way

to the end if you're Keen to carry on

the conversation with me personally I'd

love to see you over on AI ins iders on

patreon but to everyone watching have a

wonderful day