GPT Q* Strawberry Imminent, Sam Altman Trolls (Model Already Secretly Live??)

Summary

The video script discusses recent speculation around OpenAI's potential release of a new model, possibly named 'Strawberry' or 'GPT 5', which is believed to have advanced reasoning and planning capabilities. It delves into Sam Altman's cryptic tweets, the appearance of mysterious models on LM cis.org, and the community's reactions. The script also explores the potential features of 'Strawberry', including its ability to autonomously navigate the internet and perform deep research, and compares it to other AI advancements. Viewers are teased with tests of AI reasoning, hinting at the new model's capabilities while questioning if the hype is justified.

Takeaways

• 🍓 Sam Alman's tweet with a picture of a garden and strawberries fueled speculation about the possible release of the 'Strawberry' AI model, thought to be the next iteration from OpenAI.
• 🤖 Two anonymous models appeared on LM cis.org, a platform where OpenAI has previously released models, but they were not accessible to the script reader at the time of recording.
• 🕵️‍♂️ 'Jimmy Apples,' known for leaking OpenAI information, reported on a new model named 'Anonymous chatbot' which is based on the GPT-4 architecture and fine-tuned for chat interactions.
• 🧠 The 'Strawberry' model, previously known as 'Qstar' or 'QAR,' is rumored to be a significant advancement in AI, potentially enabling models to think ahead and plan, which is crucial for logic and reasoning tasks.
• 🔍 The script mentions the capability of 'Strawberry' to perform deep research and autonomous internet navigation, which are significant steps towards achieving AGI (Artificial General Intelligence).
• 📈 There's skepticism about the rumored capabilities of 'Strawberry,' with some suggesting that other labs, like Google's DeepMind, have already made strides in math reasoning, potentially reducing the advantage of OpenAI's new model.
• 🔑 'Plany the Prompter' managed to 'jailbreak' the new model, indicating that some individuals have already gained access to and tested the rumored 'Strawberry' model.
• 🤖 'Sus Column R' is another model mentioned, which appears to have a sophisticated chain of thought process, correctly answering a logic puzzle about a marble and a glass.
• 📊 The script also discusses the competitive landscape of AI development, noting that OpenAI needs to release a substantial update to maintain its position in the market.
• 🔮 There's anticipation and speculation about when 'Strawberry' will be officially announced, with some suggesting it could be imminent based on social media activity.
• 📝 The video script concludes with the reader's intention to conduct a full suite of tests on the new models to evaluate their capabilities in reasoning and logic.

Q & A

What did Sam Altman tweet on August 7th that sparked rumors about a new AI model?

Sam Altman tweeted a picture of a garden with strawberries, which led to speculations about the next big version of the Frontier Model from Open AI, often referred to as 'strawberry' or 'gp5' by the community.

What is the significance of the models appearing anonymously on LM cis.org?

The anonymous appearance of models on LM cis.org is a strategy used by Open AI for their previous iterations, suggesting that the models might be new versions or updates to existing AI models.

What is the role of Jimmy Apples in the AI community, and what did he discover about the new model?

Jimmy Apples is known as a notorious Open AI leaker. He discovered that the new model, referred to as 'anonymous chatbot,' claims to be based on the GPT-4 architecture, specifically fine-tuned for chat-based interactions.

What is the difference between 'QAR' and 'Project Strawberry' mentioned in the script?

QAR and Project Strawberry are the same; it's the renaming of a project that aims to give large language models the ability to think ahead and plan, which is considered a significant step towards achieving AGI (Artificial General Intelligence).

What are some of the rumored capabilities of 'Project Strawberry'?

Rumored capabilities of Project Strawberry include the ability to generate answers, plan to navigate the internet autonomously, perform deep research, and engage in post-training fine-tuning to optimize performance.

What is the significance of the 'Chain of Thought' in AI models?

The 'Chain of Thought' refers to a method of processing AI models that allows them to think more strategically, plan long-term, and explain their reasoning in a way that leads to higher quality outputs.

What does the acronym 'AGI' stand for, and why is it important in the context of Project Strawberry?

AGI stands for Artificial General Intelligence. It is important because Project Strawberry aims to advance towards AGI by improving reasoning, planning, and the ability to perform complex tasks.

What is the 'Arena Battle' mode in LM cis.org, and how does it relate to accessing new models?

The 'Arena Battle' mode in LM cis.org is a feature where users can interact with different AI models and vote on them. It is the only way to access the new models as they only reveal which model is being used after the user has voted.

What is the 'marble in a glass' logic problem, and why is it significant in testing AI models?

The 'marble in a glass' problem is a complex logic and reasoning test where an AI must explain the location of a marble after a series of actions. It is significant because it tests the AI's ability to understand and explain its reasoning process.

What is the correct answer to the 'marble in a glass' logic problem, and how did the AI models perform in the script?

The correct answer is that the marble would be on the table outside of the microwave. In the script, the AI models struggled with this problem, with only 'sus column R' providing the correct reasoning and answer.

Outlines

00:00

🍓 Speculations on OpenAI's 'Strawberry' Model

The video discusses the buzz around Sam Altman's tweet hinting at a garden with strawberries, which has been interpreted by the AI community as a potential reference to OpenAI's next-generation model, possibly named 'Strawberry' or 'GPT 5'. The video examines rumors and the appearance of two anonymous models on LM cis.org, which is a known strategy for OpenAI to introduce new models. The narrator explains that despite not being able to find the models, there are credible reports of their existence. It also covers the response from 'Jimmy Apples', an open AI leaker, who suggests that the new model might be based on GPT-4 architecture but does not show significant improvements in reasoning. The video further delves into the potential capabilities of 'Project Strawberry', which is believed to enhance large language models with forward-thinking and planning abilities, pushing the boundaries towards Artificial General Intelligence (AGI).

05:00

🧠 Analyzing 'Strawberry' Model's Reasoning Abilities

This paragraph focuses on the reasoning capabilities of the 'Strawberry' model, comparing it with other models like GPT 4. It highlights the importance of Chain of Thought for improving output quality in AI models. The video mentions a Twitter user, 'I Ruled the World Mo', who is considered a significant hype man for the 'Strawberry' project, posting extensively about it. The narrator also discusses the difficulty in accessing the new models on LM cis.org and shares a method to access them through Arena Battle mode. The video includes a logic problem test involving a marble and a glass, comparing the responses from GPT 4 and the new 'sus column R' model, which shows a more step-by-step reasoning approach. The paragraph concludes with the narrator's intention to conduct a full suite of tests and invites viewers to share their thoughts on the 'Strawberry' model's potential release.

Keywords

💡Sam Altman

Sam Altman is a prominent figure in the tech industry, known as the CEO of OpenAI. In the context of the video, he is discussed as a potential 'troll' due to a tweet that caused a stir in the AI community, suggesting a possible release of a new AI model, which is a central theme of the video.

💡GPT-5

GPT-5 refers to the speculated next-generation AI model from OpenAI, which is believed to be more advanced than its predecessors. The video discusses the hype and rumors surrounding GPT-5, indicating its potential capabilities and the community's anticipation for its release.

💡Strawberry

In the video, 'Strawberry' is used as a codename for a supposed new AI model from OpenAI, which is thought to be a successor to the current models. It is tied to the narrative of the video by being the subject of speculation and excitement among AI enthusiasts.

💡AI Twitter Sphere

This term refers to the community of individuals on Twitter who are interested in artificial intelligence. The video script mentions that this community 'went nuts' following a tweet by Sam Altman, highlighting the influence of such a community in spreading and discussing AI-related news.

💡LM cis.org

LM cis.org is mentioned as a platform where AI models are anonymously tested. The video discusses how two new models appeared on this site, sparking further speculation about the release of a new AI model, which is a key point in the video's exploration of AI advancements.

💡Jimmy Apples

Jimmy Apples is referred to as a 'notorious OpenAI leaker' in the script, indicating that he is known for sharing insider information about OpenAI's developments. His comments on a new model's capabilities are used in the video to provide insight into the potential features of the rumored AI model.

💡QAR

QAR, or 'Question Answering with Reasoning,' is a concept related to the ability of AI models to not only provide answers but also explain their reasoning process. The video suggests that the new model 'Strawberry' might have enhanced QAR capabilities, which is a significant aspect of the discussion around its potential features.

💡AGI

AGI stands for 'Artificial General Intelligence,' which is the idea of machines having 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. The video discusses how the rumored features of 'Strawberry' could be a step towards achieving AGI.

💡Fine-tuning

In the context of AI, fine-tuning refers to the process of further training a model on a specific task after its initial training phase. The video mentions that 'Strawberry' might involve post-training fine-tuning, which is a key feature that differentiates it from previous models.

💡Sus Column R

Sus Column R is mentioned as a model name in the script, which is part of the testing process to evaluate the reasoning capabilities of AI models. The video uses the performance of this model on certain logic puzzles to demonstrate the advancements in AI reasoning and problem-solving.

💡Chain of Thought

The term 'Chain of Thought' in the video refers to a method or technique that allows AI models to think more strategically and explain their reasoning in a step-by-step manner. It is highlighted as a potential feature of the new AI model 'Strawberry' and is used to illustrate the improvements in AI's ability to provide logical and reasoned responses.

Highlights

Sam Alman's tweet about summer in the garden sparks speculation about the release of 'gp5 strawberry', the next big AI model from OpenAI.

Two anonymous models appear on LM cis.org, indicating a potential new release from OpenAI.

Jimmy Apples, known for leaking OpenAI models, interacts with a new model named 'anonymous chatbot'.

The new model claims to be based on the GPT-4 architecture, fine-tuned for chat interactions.

Jimmy Apples notes no significant reasoning improvements but mentions potential advancements in math capabilities.

Hater atlow provides an in-depth analysis of 'Project Strawberry', suggesting it could bring planning and reasoning abilities to large language models.

Project Strawberry is believed to enable AI models to autonomously navigate the internet and perform deep research.

The concept of continuous fine-tuning and learning in AI models is presented as a significant advancement towards AGI.

Plany the Prompter claims to have 'jailbroken' an anonymous chatbot model, showcasing its capabilities.

Bendu Ready from Abacus AI suggests that other labs, including Google, have made progress in math reasoning, potentially reducing the advantage of 'Strawberry'.

A model named 'sus column R' demonstrates an advanced chain of thought in its responses to logical reasoning questions.

The term 'chain of thought' is discussed as a method to improve AI reasoning and strategic planning.

The Twitter account 'I Ruled the World Mo' is highlighted as a significant hype generator for 'Project Strawberry'.

Sean Ralston provides a method to access the new models on LM cis.org through the Arena Battle mode.

The new model 'sus column R' correctly answers a complex logic problem about a marble in a glass and a microwave.

The video concludes with a teaser for a full suite of tests on the new models and a call to action for viewer engagement.

Transcripts

Sam Alman is either an enormous troll or

gp5 strawberry is right around the

corner let's break down all of the

rumors and the hype that have been

really building over the last couple

days so just today as of recording this

video on August 7th 829 a.m. Pacific Sam

ultman tweets out I love summer in the

garden what a troll he took a picture of

a garden with strawberries and if you're

not familiar strawberry or qar or

whatever you want to call it is what

everybody thinks is the next big version

the next Frontier Model from open Ai and

of course after this tweet the AI

Twitter sphere went nuts and everybody

started commenting on it but that's not

it there were actually two Anonymous

models that just appeared in LM cis.org

this is the same strategy that open AI

used for their previous iterations of

models just anonymously dropping the

models in LM cis.org now I went there

this morning and I could not find either

of these models but there's been enough

reports throughout the internet people

that I hopefully can trust that have

showed these models in action so here is

Jimmy apples the notorious open AI

leaker new model in lmis Arena Battle

only as we can see here the model name

is anonymous chatbot now for other

people it's showing up as a different

name which means it might actually be a

completely different model we'll get to

that so here he asked what model are you

on based on opena I GPT 4 architecture

specifically you're interacting with a

version of gp4 that has been fine-tuned

for chat based interactions blah blah

blah so not much but it is saying it is

GPT for architecture but who knows if

that's true or false Jimmy apples goes

on to say from some very rough and

limited personal testing I'm not seeing

any reasoning improvements but I've seen

some in math maybe someone with better

personal evals on math can test it I

wish I could test it I cannot find it

anywhere in lm.org then we have hater

atlow developer who broke down

everything we know about qar strawberry

so let me just quickly go over this so

it's confirmed open AI is close to

announcing its next Frontier Model

possibly GPT 5 open AI has renamed

project qstar to project strawberry and

for those of you who are asking what is

Project strawberry what is qar I've made

multiple videos about them in the past

the gist is it is finally giving large

language models the ability to think

ahead to plan which allows them to get

better at math to get better at logic

and reasoning and really is an enormous

unlock towards AGI if true so there's

been a bunch of rumors about qar about

strawberry and here are just a few of

what people think it might be capable of

it will generate answers but also plan

enough to navigate the internet

autonomously and reliably to perform

deep research deep research planning

actually being able to think through a

prompt rather than just immediately

responding with whatever it is trained

on it involves a specialized way of

processing an AI model after it has been

pre-trained trained on large data sets

so typically how it works is a model is

initially trained and then it's kind of

Frozen in time until it's fine-tuned

later but this idea that it can be

consistently fine-tuned and consistently

learning rather than just a knowledge

based Frozen in time is an incredible

and Elusive idea in the world of AI so

some key points this reasoning is key to

AGI and Asi open AI wants models to

browse the web with the assistance of

computer using agents and take actions

based on their findings they want

Strawberry to perform long Horizon tasks

to perform a series of actions over an

extended period and this is something

Sam Alman has talked about in previous

interviews it will engage in

posttraining fine-tuning that optimizes

performance after the regular training

phase and of course plany the prompter

got his or her hands on this model so

Model A Anonymous chatbot and was

already able to jailbreak it plany the

prompter is ruthless however bendu ready

from Abacus AI has a slightly different

take on it but yes this is in reference

to project strawberry qar the reasoning

project open AI has been rumored to be

working on the problem however is that

several other labs including Google have

cracked a bunch of techniques around

math reasoning and synthetic data now

what she is specifically referring to is

just about a week ago deep Minds Model A

company that is owned by Google was able

to absolutely dominate the math

olympiads so basically the whole math

reasoning thing is nearly solved so she

goes on to say it's unlikely that

strawberry is going to give them much

advantage over Opus 3.5 or Gemini 2.0

now here's that other model I was

telling you about this is a screenshot

from a DTS Singh let's take a look the

model is sus column R what a name so

here's one of the questions that has

been going around the internet I've been

including it in my llm test and the

question is which is larger 9.11 or 9.9

and not only did it give the answer but

it also gave the reasoning as to how it

arrived at the answer and it did give

the correct answer but a lot of models

have been struggling with this very

simple prompt and the DT says sus colr

seems to have insane coot Chain of

Thought built in maybe qar chubby also

somebody who is a great follow on

Twitter says why Chain of Thought and

not tree of thought so they're really

talking about algorithms or really just

prompting techniques to allow models to

think more strategically to think more

longterm to plan and to really explain

their reasoning in a way that allows

them to have much better quality outputs

and I can't end this video about

strawberry with talking about I ruled

the world Mo and this is a newish

Twitter account at least new to me and

is possibly the biggest troll the

biggest hype man for strawberry qar

there possibly is I don't know who he is

I think he's actually just an ALT of

chubby but maybe not maybe he's an

Insider at open AI who do you think he

is by the way drop your comments in the

description below and maybe he'll show

up in the comments and reveal himself

but he has already 8,800 Plus posts

which is just insane to think about and

look at some of these posts choo choo

project strawberry and on and on and on

all about project strawberry all about

hyping it up and we'll see if it's

actually true if project strawberry is

coming tonight which I rule the world

says it is or if it's coming soon you

know open AI is at the point where they

really have to drop something

substantial very soon because llama 3.1

405b took a lot of the wind out of open

AI sales well it turns out right after I

finished recording Sean rousson actually

told me how to get access to these

models there is a reason I couldn't find

them they only show up in the arena mode

battle mode of LM cis.org and it doesn't

tell you which model is actually being

used until after you vote on it so the

only way to get access to it is to

basically try a bunch of prompts in the

Arena Battle and then hopefully you get

the new model and I did check this out

so I asked the Killer's question there

are three killers in the room someone

aners the room and kills one of them

nobody leaves the room how many killers

are left in the room and we have two

models GPT

40613 and here it is sus column R now

let's look at GPT 4 first let's break it

down initially there are three killers

in the room someone else enters the room

and kills one of the existing three

killers nobody leaves the room here's a

critical part of the answer the person

who entered the room and killed one of

the killers becomes a killer so there

are still three killers in the room okay

great now let's look at this new model

so initial situation action result of

the action additional consideration

final count so it's definitely breaking

it down in a much more stepbystep

approach so we get therefore there are

two original Killers plus one new killer

equals three killers and it actually

gives us both answers three killers if

you include the person who committed the

recent killing and two killers if you

only count the original ones now still

it doesn't give any note about the

killer who is now dead and it probably

should have but let's give it one more

now here's the hardest logic and

reasoning problem I have in my llm tests

a marble is put in a glass the glass is

then turned upside down and put on a

table then the glass is picked up and

put in a microwave where is the marble

explain your reasoning step by step so

we don't actually know which models are

generating these responses yet so let's

look at model A the marble is inside the

glass which is inside the microwave that

is dead wrong however let's look at

model B step one the Marble's put in the

glass the glass is then turned upside

down the glass is then picked up the

marble Still Remains ins inside the

glass unaffected by the change in

location and they both say it is in the

microwave so both are bad and yep

neither of these are the new models

let's try it again so I did it again and

we have a new model this is anonymous

chatbot what everybody is also saying is

potentially the next version of GPT

however it does say the marble is at the

bottom of the glass when the glass is

placed in the microwave which is wrong

that is just wrong so it did get this

wrong and I want to thank again Sean

Ralston for sending me this rather than

just testing it a bunch of of times and

trying to get the right model we have

the marble test from sus column R let's

see what it says initially the Marble's

placed inside a normal cup the cup is

then turned upside down over the marble

due to gravity the marble would fall to

the bottom of the inside of the cup

positioning under the cup when the cup

is upside down the marble would rest on

the inner surface of the Cup's bottom

due to gravity yes but it's confined by

the Cup's walls and the table surface

yes moving to the microwave the cup is

lifted the marble will no longer be

supported by the Cup's interior and will

fall due to gravity it will now be lying

on the table where the cup was yes yes

the marble would now be located on the

table outside of the microwave that is

correct now of course I plan on doing

the full Suite of test but I just wanted

to show you a couple of the really hard

reasoning tests before I actually do the

full test in a separate video let me

know if you think strawberry is coming

soon is Sam Alman trolling we'll see if

you enjoyed this video please consider

giving a like And subscribe and I'll see

you in the next one