The AI Hype is OVER! Have LLMs Peaked?

Summary

The video script discusses the current perception of AI, particularly generative AI and large language models (LLMs), and counters the notion that AI hype is fading. It references the Gartner Hype Cycle to illustrate the stages of technological maturity and adoption, arguing that while some believe we are at a plateau, significant advancements are still to come. The speaker highlights key developments such as OpenAI's Sora, advancements in voice recognition, and the potential for AI to exceed human performance in complex tasks. They also emphasize the importance of energy and compute capacity as potential bottlenecks for future AI growth. The script suggests that internal developments at companies like OpenAI are far from slowing down, with暗示 (hints) at groundbreaking models like GPT 5 that could revolutionize the field. The summary concludes by emphasizing that despite external appearances, the AI industry is on the cusp of significant breakthroughs that will propel it into a new era of innovation and capability.

Takeaways

• 🚀 **AI Hype and Progression**: Despite perceptions that AI hype is waning, significant advancements are still being made in the field of generative AI, with large language models (LLMs) continuing to evolve.
• 📈 **Gartner Hype Cycle**: The Gartner Hype Cycle is a useful tool for understanding the trajectory of technological development, but some argue that AI, particularly LLMs, is not following the traditional pattern of the cycle.
• 💡 **Innovation in AI**: Companies like OpenAI are making consistent breakthroughs in AI, some of which are not publicly disclosed, indicating that internal progress is much faster than what is visible externally.
• 🔥 **Energy and Compute Bottlenecks**: Energy consumption and compute capacity are significant bottlenecks for the future of AI, with the need for vast amounts of power and advanced infrastructure to support increasingly complex models.
• 💼 **Business Dynamics**: The shift from open research to closed, proprietary development means that companies are less likely to share their innovations, leading to a perception that progress is slower than it actually is.
• 🌐 **Global AI Race**: There is a global race in AI development, with companies investing heavily to stay competitive, which is driving rapid advancements and the potential for new breakthroughs.
• 🔬 **Advanced Reasoning Engines**: The incorporation of advanced reasoning engines on top of existing models like GPT-4 is pushing the boundaries of what AI can achieve, particularly in complex problem-solving tasks.
• 📊 **Benchmarking and Competition**: Companies are incentivized to surpass the current benchmarks like GPT-4, leading to a focus on marginal improvements rather than acknowledging the potential for more significant leaps in capability.
• 🔍 **Internal Developments at OpenAI**: There is an expectation that OpenAI is significantly ahead of its current public releases, suggesting that future releases could bring substantial advancements in AI capabilities.
• 🔥 **Sora and Other Innovations**: OpenAI's release of Sora, a video generation model, and other undisclosed projects indicate that the company is exploring various AI applications beyond just text-based models.
• ⏳ **Future Predictions**: Industry leaders like Sam Altman suggest that the current state-of-the-art models will be considered primitive in the near future, hinting at the potential for massive jumps in AI performance.

Q & A

What is the current debate about AI on social media platforms like Twitter?

The current debate is centered around whether the hype surrounding AI, particularly generative AI, is wearing off and if we have reached an exhaustion point in terms of its capabilities and future developments.

What is Sam Altman's stance on the investment in building AGI, despite the high costs?

Sam Altman is quoted as saying that he doesn't care if they burn $50 billion a year because they are building AGI, and he believes it will be worth it.

What is the Gartner Hype Cycle and how is it relevant to AI technologies?

The Gartner Hype Cycle is a graphical representation used to illustrate the maturity, adoption, and social application of specific technologies. It provides a conceptual framework to understand how technologies evolve from introduction to mainstream adoption, which is particularly useful for evaluating emerging technologies like generative AI and large language models.

What are some of the current criticisms of large language models (LLMs)?

Criticisms include biases in the models, the need for vast data training, high operational and inference costs, environmental impacts, and issues like hallucination where the model generates incorrect or misleading information.

What is the 'slope of Enlightenment' in the Gartner Hype Cycle?

The 'slope of Enlightenment' is a phase where previous issues with a technology are ironed out. For LLMs, this would involve improvements in model efficiency, reduced biases, and increased reliability.

What are some of the advancements in AI that suggest it is not stagnating?

null

What does Sam Altman imply about the future of AI models beyond GPT-4?

Sam Altman implies that future models, such as GPT-5, will represent a significant leap from the current state-of-the-art, suggesting that GPT-4 will be considered 'dumb' in comparison to what is to come.

What are the potential bottlenecks that could slow down the progress of generative AI?

The potential bottlenecks include energy constraints, due to the high power requirements for training and running large models, and compute capacity, as the demand for processing power may exceed current supply.

null

null

Why might companies be hesitant to publicly release all their AI research and advancements?

Companies like OpenAI may keep their research and advancements secret to maintain a competitive edge. Revealing too much could allow competitors to replicate their successes and potentially surpass them.

How does the release of GPT-4 as a benchmark affect the AI industry?

The release of GPT-4 as a benchmark creates an incentive for other companies to train models that just surpass GPT-4, rather than aiming for a significant leap in capabilities. This can create an illusion of a plateau in AI progress.

What does the future hold for AI capabilities according to the advancements and statements made by industry leaders?

The future of AI is expected to be transformative, with significant advancements in reasoning capabilities, efficiency, and the potential for AI agents to perform complex tasks at a level surpassing current human capabilities.

Outlines

00:00

🤖 AI Hype and the Gartner Hype Cycle

The first paragraph discusses the perception that the hype around AI might be fading, particularly in the context of generative AI. It references a clip from Sam Altman, emphasizing the commitment to developing AGI despite the costs. The paragraph also introduces the Gartner Hype Cycle as a tool for understanding the trajectory of technological development, from introduction to mainstream adoption. It outlines the cycle's stages, including the technology trigger, peak of inflated expectations, and the trough of disillusionment, which is where many believe generative AI currently stands due to issues like biases, high costs, and environmental impact.

05:02

🚀 The Upcoming Surge in AI Advancements

This paragraph challenges the notion that AI is stagnating and argues that we are on the cusp of significant breakthroughs. It points out that while some benchmarks show AI nearing human performance, there are still complex tasks where AI lags, suggesting room for growth. The speaker disagrees with the idea that we have reached the limits of generative AI and alludes to ongoing advancements in various AI domains, such as voice recognition and video generation, indicating that the field is far from plateauing.

10:03

📉 The Downturn in AI Hype and Future Predictions

The third paragraph focuses on the idea that despite the perceived downturn in AI hype, industry leaders hint at upcoming revolutionary AI models. Sam Altman's statements suggest that current state-of-the-art models like GPT-4 are considered 'dumb' compared to what's coming next, indicating a significant leap in AI capabilities. The paragraph also discusses the potential for energy consumption and GPU production to become bottlenecks in AI development, highlighting the infrastructure challenges that could slow progress.

15:06

💡 The Impact of Energy and Compute Constraints on AI

This paragraph delves into the challenges of energy consumption and the supply constraints of GPUs, which are critical for training large language models. It discusses the potential for energy to be a limiting factor in AI's growth, given the extensive power requirements for training clusters. It also touches on the regulatory hurdles and long lead times associated with building new power plants and infrastructure, suggesting that these factors could slow the pace of AI advancement.

20:07

🌐 The Unseen Progress in AI and the Shift to Closed Research

The fifth paragraph emphasizes that while external appearances may suggest a slowdown in AI progress, internal developments at companies like OpenAI are likely far more advanced. It highlights the shift from open research to closed, proprietary research as companies protect their innovations. The paragraph also mentions the competitive drive among companies to surpass GPT-4, suggesting that the current benchmark does not indicate a plateau but rather a strategic pause before the next significant release.

25:07

🔮 Future AI Capabilities and the Race for Advancements

The final paragraph speculates on the potential for AI to surpass current benchmarks significantly, leading to transformative changes across industries. It discusses the potential integration of advanced reasoning engines with current models like GPT-4 and the impact of iterative agent workflows. The speaker expresses optimism about the unseen advancements within companies like OpenAI and predicts that the next year will bring groundbreaking developments in AI.

Keywords

💡AI Hype

AI Hype refers to the heightened public interest and media coverage surrounding artificial intelligence technologies, particularly generative AI and large language models. In the video, the speaker discusses the perception that the excitement around AI might be fading, but argues against this notion, suggesting that significant advancements are still to come.

💡Generative AI

Generative AI is a branch of artificial intelligence that involves creating new content, such as text, images, or videos, that did not exist before. It is a key focus of the video, with the speaker highlighting how this technology is evolving and disputing claims that its development is slowing down.

💡Gartner Hype Cycle

The Gartner Hype Cycle is a graphical representation that illustrates the life cycle of a technology, from its initial emergence to eventual mainstream adoption. The video uses this concept to discuss the stages of technology adoption, particularly in relation to AI, and to argue that AI is not yet at its peak in terms of public interest and capability.

💡Large Language Models (LLMs)

Large Language Models (LLMs) are AI systems that process and generate human-like language at scale. They are a central topic in the video, where the speaker discusses their current capabilities, the challenges they face, and the potential for future advancements that will significantly surpass current models.

💡Sam Altman

Sam Altman is the CEO of OpenAI and a key figure in the development of AI technologies. In the video, his statements about the future of AI and the willingness to invest heavily in AGI (Artificial General Intelligence) are used to illustrate the commitment and vision of industry leaders in pushing the boundaries of AI.

💡AI Efficiency and Inference Costs

AI Efficiency and Inference Costs refer to how well AI models perform with the resources they use and the expenses associated with running them. The video discusses these costs as a challenge for AI development, highlighting the need for more efficient models that reduce operational and environmental impacts.

💡Bias in AI

Bias in AI refers to the systemic errors or unfairness that can occur in AI systems due to the data they are trained on or the algorithms they use. The video touches on this issue as one of the problems that need to be addressed as AI technology matures, emphasizing the importance of reducing biases for more reliable and fair AI systems.

💡Energy Constraints

Energy Constraints highlight the limitations and challenges posed by the high energy consumption of AI systems, particularly when it comes to training large models. The video suggests that energy costs and infrastructure could become a bottleneck in the rapid advancement of AI, affecting the pace at which new models are developed and deployed.

💡Compute Capacity

Compute Capacity refers to the ability of hardware to perform computational tasks, which is crucial for training and running AI models. The video discusses the need for increased compute capacity to meet the demands of future AI systems, with references to the development of supercomputers to handle the complex tasks of advanced AI.

💡Closed Research Environment

A Closed Research Environment is a situation where research and development are conducted privately within a company, with findings not shared publicly. The video suggests that companies like OpenAI have moved from an open research model to a closed one, keeping their innovations secret to maintain a competitive edge and not give away their advancements to competitors.

💡Multimodal AI Agents

Multimodal AI Agents are systems capable of processing and understanding multiple types of data inputs, such as text, images, and audio. The video speaks to the potential of these agents in performing complex tasks and the significant leap forward this represents for AI capabilities, suggesting that when these agents reach a certain level of performance, it will mark a transformative moment for AI.

Highlights

AI hype is not wearing off; generative AI's capabilities are still growing.

Sam Altman's statement about investing heavily in AGI development signifies ongoing commitment to AI advancement.

Generative AI is not slowing down; it's evolving with new developments in voice recognition and video generation.

Gartner hype cycle is used to illustrate the maturity, adoption, and application of technologies like generative AI.

The peak of inflated expectations for LLMs was reached with GPT 3.5 and GPT 4, indicating significant media attention and public interest.

The trial of disillusionment phase is characterized by the recognition of issues like biases, high operational costs, and environmental impacts of large language models.

The slope of Enlightenment suggests that LLMs are improving in efficiency, reducing biases, and increasing reliability.

AI performance on benchmarks is approaching human levels in areas like image classification and natural language inference, but complex tasks remain a challenge.

Sam Altman hints at significant advancements beyond GPT 4, suggesting that future models will be dramatically more capable.

Energy consumption and GPU supply constraints are potential bottlenecks for the growth of AI systems.

NVIDIA's Blackwell GPU architecture promises significant performance improvements for training large language models.

OpenAI's shift from open research to closed research environment means they are making consistent breakthroughs without immediate public disclosure.

GPT 4 as a benchmark creates an illusion of a plateau, as companies aim to just surpass it before releasing their models.

The incorporation of advanced reasoning engines on top of existing models like GPT 4 is pushing the boundaries of AI capabilities.

Andrew NG's research on iterative agent workflows shows significant improvements in AI performance, indicating ongoing progress in the field.

Internal advancements at companies like OpenAI suggest that there will be shocking releases in the future as they advance beyond current benchmarks.

The next 365 days are predicted to bring about significant changes and advancements in AI, particularly in multimodal agents and open-ended tasks.

Transcripts

so one of the current questions that

I've been seeing floating around on

Twitter and on social media is that AI

hype might be wearing off and when I'm

talking about AI hype this is referring

to the trend that generative AI has I

guess you could say reached an

exhaustion point now I disagree and I'm

going to tell you guys why but I'm going

to go over some of the key examples and

what people are stating so essentially

there was a clip here and a lot of

people were stating that this clip from

Sam Altman people were stating that the

hype is wearing off The Vibes are

shifting you can feel it and basically

in this clip clip here Sam Alman

literally is he's only really stating

that you know I don't care if we burn

$50 billion a year we're building AGI

and it's going to be worth it so it's

not a crazy crazy statement I think why

people are stating that the hype is

wearing off and that Vibes are wearing

off and this video is going to be and

essentially what this is referring to

here is the fact that you know I guess a

lot of people are thinking that you know

I guess you could say the generative AI

is currently slowing down in terms of

the capabilities and what we're likely

to see in the future now like I said

before it's just mainly due to this clip

and a bunch of other different factors

but let's get into how Cycles work in

terms of technology and a graph that a

lot of people have been referencing when

they talk about llms plateauing so

something that I also saw quite a bit

being you know passed around as like

this infographic is of course the

Gartner hype cycle so essentially it's

just a graphical rep representation used

to illustrate maturity adoption and

social application of specific

Technologies it basically just provides

a conceptual framework that helps

stakeholders and individuals understand

how Technologies evolve from the time

they introduced until they reach

mainstream adoption and the hype cycle

can be particularly useful for

evaluating emerging Technologies like

generative Ai and of course large

language models so of course we

essentially have the technology trigger

this phase initially occurs when a new

technology is conceptualized or when a

significant development makes it

publicly known creating interest and

significant media attention for example

this is when GPT 3.5 chat GPT was

released and it demonstrated the ability

to create text in the long

format and then of course that's when we

got to GPT 4 which is stage two so this

is where there is the peak of inflated

expectations now I've got to be honest

it's not just llms that were going crazy

at this point there was also 11 labs and

other image generation services like mid

journey and of course other services

like stable diffusion so I would argue

that the problem with this is that with

generative AI experiencing some kind of

increase in the actual media coverage I

would say that this is something that is

cumulatively increasing in terms of the

expectations and that's because like I

said already there were many several

different categories that came together

at the same time and of course this

could include llms being at the peak of

inflated expectations now I want to say

that I do disagree with this hype cycle

for AI I do think that this is nowhere

near it its peak where it should be but

there of course is inflated expectations

when a new technology comes to fruition

a lot of people may exaggerate what the

technology can really do for example

some people say it's going to replate

replace entire careers or entire tasks

and revolutionize entire Industries now

whilst yes that might happen in the

future I don't think that that is

completely happening with GPT 4 and gbt

3.5 so some are arguing that this is

where we are at currently okay and then

of course the trial of disillusionment

and this is where Technologies enter

this phase when they fail to meet the

inflated expectations and stakeholders

become increasingly delusion so issues

related to the technology start to arise

such as the biases and large language

models need for vast data training the

high operational costs the high

inference costs the environmental

impacts they become more apparent and

then they become criticized and one of

the main things that many people are

talking about with llms is of course

things like the hallucination and the

high inference costs and of course the

training cost because these models are

certainly not cheap and it seems like

there's only a few companies that can

run and train these large models now of

course there we have the slope of

Enlightenment and this is essentially

where with l M all of these previous

issues get ironed out so things like the

issues of hallucination and the biases

they get ironed out right here and this

is where as more experimentations as

more implementations occur the market

matures and second third generation

products appear and with llms this would

involve developments that address the

earlier criticisms such as improving the

model efficiency the inference costs

reducing the biases and of course the

reliability of the model and this is

arguably where people think AI is going

to go technology becomes stable and

accepted and this means widespread

adoption however I think the graph isn't

going to look anything like that I think

it's probably going to look something

like this where we go up we dip a little

bit and then we continue a trajectory

upwards because like I said before

whilst yes that it does seem to many

different pieces of Statistics that it

looks like we are slowing down in terms

of AI versus human performance and even

on this Stanford AI index we can see

that AI has surpassed human performance

on several benchmarks including someon

image classification visual reasoning

and English understanding yet it Trails

behind on more complex tasks like

competition level mathematics visual

Common Sense reasoning and planning and

you can see here that if we take a look

at actually how the AI is moving we can

see that in image classification visual

Common Sense reasoning natural language

inference all of these seem to be coming

towards the human level Baseline but

they don't seem to be going upwards you

know like in a crazy level on on the

graph and some people would argue that

this is because AI generative AI large

language models whatever you want to

call it if you just want to group

everything together that we have reached

our limit in terms of where we are and

new architectures are going to be needed

now I would firmly disagree with this

for several reasons that I'm about to

explain and I'm going to show you guys

some really key evidence on why things

are about to actually get very very

crazy in the world of AI and why we're

about to go into a very very abundant

era due to AI so one of the things that

many people are not actually taking into

account is the fact that currently there

are many different things going on in

the world of AI that people aren't

paying attention to some people are just

paying attention to large language

models and this doesn't make sense

because there are vast and many

different categories in which AI is

currently exceeding for example in voice

recognition and in voice generation

opening ey has recently developed their

voice engine which was actually

something that they developed in around

2022 and it was basically a

state-of-the-art system that could

recreate anyone's voice in addition to

that if you do remember open ey also did

talk about and quote unquote release

their sore up model to Showcase us how

far they've come in creating a text

video model now I think you have to

understand how crazy this is because for

people to say that generative AI is

stagnating is a crazy statement when

literally this year we literally got

Sora which blew everyone's Minds in

terms of the capabilities so with Sora I

still find it absolutely incredible that

people can say AI is stagnating because

with Sora as you've all seen this was

something that was truly just

mind-blowing this piece of technology

showed us how crazy it is when you get a

group of AI researchers dedicated to

doing something and I'm going to give

you guys a quick memory joke remember

that opening ey isn't a company that's

Focus fed on video generation this is

just a subsection of their company so

the video generation aspect was

something that I guess they just wanted

to see if they could do well at and they

literally surpassed state-of-the-art

models Google Runway pabs they

completely surpassed them and this is

absolutely crazy I mean the demo

literally absolutely shocked everyone I

was truly truly speechless when this

technology was here and I'm someone that

pays attention to all of the AI news and

of course we do have Devon cognition

Labs released Devon their first AI

software engineer and this was a AI

rapper around GPT 4 but it did a few

things in a unique way where it was able

to surpass GPT 4 on certain software

engineering benchmarks so one of the

first things the point here I'm making

is that you might think that llms are

stagnating but that is not the truth at

all we're going to get into llms later

but overall voice engine Sora and Devon

show us that generative AI is really

really not going to be stagnating

anytime soon but if that didn't convince

you let me show you guys some of the

recent statements that show you that

we're absolutely in for a crazy ride so

Sam Alman recently said in an interview

at Stanford's entrepreneurship talk he

spoke about gp4 now remember GPT 4

currently is a state-of-the-art system

meaning that it is the best of the best

that we can currently get our hands on

for public use which means that Sam

Alman currently probably has access to

Frontier that are being developed by

open Ai and remember the big Labs like

anthropic and Google are currently

behind them in terms of what they're

creating so take a listen to this

statement open ey is phenomenal chat gbt

is phenomenal um everything else all the

other models are phenomenal it burned

you've burned $520 million of cash last

year that doesn't concern you in terms

of thinking about the economic model of

how do you actually where's going to be

the monetization source well first of

all that's nice of you to say but Chachi

BT is not phenomenal like Chachi BT is

mildly embarrassing in a best um gp4 is

the dumbest model any of you will ever

ever have to use again by a lot um but

you know it's like important to ship

early and often so if you weren't paying

attention there Sam mman literally just

said that this is going to be the

dumbest model that we will have to use

or that we will have had to use by far

so he didn't just state that this was

going to be an incremental increase he

clearly stated that gp4 was dumb he

stated this model was not you know great

he stated that this model was not that

good he clearly stated that GPT 4 a

current state-ofthe-art system that

people were literally able to get you

know increasingly capabilities just by

wrapping the system and being able to do

software engineering tasks and a lot of

people even using GPT 4 to be able to

train robotic systems like recently we

saw in a research paper and this was

literally where we had language model

guided Sim to real transfer so we

basically had large language models were

basically writing the reward functions

for an AI system and it was able to do

it very effectively and it literally

just went from simulation to real life

which means that this is going to

immediately speed up how quickly we're

able to train robots and have them you

know doing well in novel environments

it's pretty pretty insane so it's an llm

guided Sim toal approach so that is

pretty crazy and remember I'm guessing

that they were using GPT 4 that's

actually a study that I haven't covered

just yet but the point is is that samman

states that the current state-of-the-art

model the current model that other

industry labs are trying to be millions

and millions of dollars he states that

that model is bad and it's dumb he

didn't just say it was kind of smart he

said that it was dumb which leads me to

believe that we are truly truly not even

scratching the surface for what AI

systems are he could have just said

future models will be interesting he

could have just said they will be you

know kind of good but he literally said

okay I'm going to play it one more time

first of all that's nice of you to say

but Chachi BT is not phenomenal like

Chachi PT is mildly embarrassing at best

um GPT 4 is the dumbest model any of you

will ever ever have to use again Again

by a lot um but you know it's like

important to ship early and often so

that just goes to show how clear there's

going to be a distinction in the future

now he also stated that there will be a

massive jump from DPT 3.5 to 4 and there

will be a similar jump from GPT 4 to GPT

5 so it's important to know the jump

from GPT 3.5 to GPT 4 was incredible

because of GPT 3.5 limitations it meant

that it couldn't be extended to certain

tasks but if we have that same jump from

GPT 4 to GPT 5 then things are truly

about to change there's so many

questions uh first of all also amazing

it's looking back it'll probably be this

kind of historic pivotal moment with 35

and four which had your BT maybe five

will be the pivotal moment I don't know

hard to say that looking forwards we

never know that's the annoying thing

about the future it's hard to predict

but for me looking back GPT 4 Chad GPT

is pretty damn impressive like

historically impressive so allow me uh

to ask what's been the most impressive

capabilities of GPT 4 to you and gp4

turbo I think it kind of sucks H typical

human also gotten used to an awesome

thing no I think it is an amazing thing

um but relative to where we need to get

to and where I believe we will get to uh

you know at the time of like gpt3 people

were like oh this is amazing this is

this like Marvel of technology and it is

it was uh you know now we have gp4 and

look at GB3 and you're like that's

unimaginably horrible um I expect that

the Delta between 5 and four will be the

same as between four and three and I

think it is our job to live a few years

in the future and remember that the

tools we have now are going to kind of

suck looking backwards at them so you

can clearly see here that he's basically

stating that a year from now when we

have gbt 5 or the next level of Frontier

Model is released we're going to look

back at GPT 4 and think it's pretty

pretty bad so as much as people are

stating that AI hype is wearing off Sam

alman's feeling it there's this Gartner

hype cycle it's important to remember

the subtle cues the subtle tells the

subtle statements that we've clearly

seen from industry leaders about these

future future models that only they

currently have access to now if

generative AI was actually slowing down

it would be due to a bottleneck and

there are only a couple bottlenecks that

are really the issues okay one of the

issues is of course energy and this one

isn't really talked about enough because

energy isn't something that people think

of but trust me when I say these

inference costs are really really

expensive and a lot of people you know

including Mark Zuckerberg are basically

stating that one of the major

limitations for future AI systems and

where things might actually start to

slow down is due to the energy cost the

rampant energy cost for these AI systems

and just how much electricity they

consume and there were even a few rumors

talking about a GPT 6 training cluster

project that would arguably shut down

the power grid something along those

lines I know it does sound crazy and I

know rumors are rumors but energy is

expensive and a lot of it is required to

run inference on these large language

models which is why they often restrict

us I think there was this issue of um

GPU production yeah right so even

companies that had the money to pay for

the gpus um couldn't necessarily get as

many as they wanted because there was

there were all these Supply constraints

now I think that's sort of getting less

so now I think you're seeing a bunch of

companies

think about wow we should just like

really invest a lot of money in building

out these things and I think that will

go for um for some period of time there

is a capital question of like okay at

what point does it stop being worth it

to put the capital in but I actually

think before we hit that you're going to

run into energy constraints right

because I just I mean I don't think

anyone's built a gigawatt single

training cluster yet I me just to I

guess put this in perspective I think a

gigawatt it's like around the size of

like a meaningful nuclear power plant

only going towards training a model and

then you run into these things that just

end up being slower in the world like

getting energy permitted is like a very

heavily regulated government function

and if you're talking about building

large new power plants or large build

outs and then building transmission

lines that cross other private or public

land that is just a heavily regulated

thing so you're talking about many years

of lead time so if we wanted to stand up

just some like massive facility um to

power that I I think that that is that's

that's a very long-term project so

basically what Mark Zuckerberg is

stating here is that energy is going to

be a huge bottleneck because unlike

software where you can make it more

efficient and you can do things quickly

or you can get a GPU produced quickly

trying to build a nuclear power plant it

takes time trying to build this these

kind of infrastructure it's not

something you could just do in a day and

it's something that's quite important

when it comes to ative AI because as the

race continues they're going to be

spending a lot more money they're going

to be spending a lot more in terms of

how much they're spending on cooling

because these gpus heat up quite a bit

so I think this is something that you

know this is probably where the

constraints actually will come from in

the future and this is where things

might actually slow down a little bit

because there are going to be a few

things that make this truly hard to

continue with I think we would probably

build out bigger clusters than we

currently can if we could get the energy

to do it so I think that that's um

that's fundamentally money bottl in the

limit like if you had a trillion dollar

I think it's time right um but it

depends on how far the the exponential

curves go right like I think a number of

companies are working on you know right

now I think you know like a lot of data

centers are on the order of 50 megawatts

or 100 megawatts or like a big one might

be50 megawatts okay so you take a whole

Data Center and you fill it up with just

all the stuff that you need to do for

training and you build the biggest

cluster you can I think you're that's

kind of I think a bunch of companies are

running at stuff like that so it will be

interesting to see where gen of AI does

fall down but like I've said it's not

slowing down anytime soon and if you

remember open a and Microsoft are

building a100 billion Stargate AI

supercomputer to power the AGI or ASI

now there's also another bottleneck

which is what I've titled here which is

the compute problem essentially this

just means that the compute capacity for

AI systems is far too great and it kind

of exceeds the demand that we require so

that's why they're building out this1

billion supercomputer to meet the

demands of future generative AI systems

or to essentially power the next

Industrial Revolution because my oh my

if an AGI or air size here it's going to

be used pretty much everywhere to power

the economy and you're going to need the

compute and the infrastructure to do

that and currently this is one of our

biggest things because openi and

Microsoft don't even have enough chips

and don't even have enough computes

currently to compete with the likes of

Google so what we have here is we have

updates to the chips and you can see

that nvidia's recent Blackwell is pretty

pretty incredible and this was one of

the most important developments for AI

because this accelerates the training of

large language models and generative AI

systems so the Blackwell GPU

architecture with its 208 billion

transistors and its enhanced Transformer

engine is designed to dramatically

increase increase the training of large

language models like GPT 4 according to

Nvidia Blackwell can provide up to 30

times higher performance for generative

AI inference compared to the previous

h100 gpus four times faster training

performance for large language models

and this essentially means large

language models like GPT 4 which took

around 90 days to train on 8, h100 gpus

consuming 15 megawatt of power could

potentially be just trained in just 30

days on two 2,000 black C gpus only

using 4 megawatt which is pretty pretty

incredible and that curve definitely

reminds me of some that we've seen

before where things are starting to

exponentially increase so the point

right here is that these are the actual

bottlenecks of generative AI because a

lot of people are thinking that you know

things are slowing down things are

getting worse and trust me guys if you

been paying attention things I would

argue are actually speeding up

internally and one of the things that I

didn't even include in this presentation

because I forgot about it is the fact

that right now it's like open ey lit a

match under every other company because

now other companies are realizing that

whoa there's a huge huge AI race going

on and if we partake we could definitely

be getting billions and billions of

dollars and that means that other

companies and other startups are all

rushing down the corridor to see if they

can get piece of the piie which means

that we're about to see a complete

Revolution and a complete new industry

in terms of all of these products and

services now one of the biggest things

that I think that most people need to

consider is that open AI are no longer

open okay and if there's one thing you

take away from this video please

understand this okay things might be

slowing down externally but things are

not slowing down internally and what I

mean by that statement is that currently

we're at a stage where things have moved

from an open research environment to a

closed research environment the reason

this has happened is because opening up

they're no longer essentially a company

that's just focused on Research they are

a business and businesses you know they

hide their secrets and they hide their

Innovations because they don't want

their competitors to have them if open I

shared all their secrets then other

companies could easily build gbt 4 with

remarkable accuracy and op essentially

has secret Source the thing is openai

also doesn't publicize their research

I'm sure breakthroughs are made every

single month okay and you have to think

about it if openai did their whatever

they did with gbt so long ago they must

have some secret kind of breakthrough

they must have some secret source and

they must have something that others

don't which essentially means that open

ey are making consistent breakthroughs

and remember Sora they had Sora we had

no news no indication that they were

even developing some video AI there was

literally no indication whatsoever there

was no interview from Sam Alman there

was literally nothing we could have

picked up on on the fact that they were

even training such a model and boom they

just you know put it out into the open

the point is is that we know we have no

idea on what's going on at you know

closed AI open AI whatever you want to

call it the point is is that internally

I can guarantee you guys they are like 2

to 3 years ahead from where they are and

the point there is that whilst you might

think ah they haven't released anything

in a while that doesn't mean things are

slowing down it just means that they're

thinking okay how can we not Shock the

public with this next release that we

know is literally going to take

everything remember other companies are

still playing catchup gbt 4 finished

training in August of 2022 which means

that we are very very lucky because

we're going to be in for a real surprise

when gbt 5 gets here now another thing

to note as well okay is that GPT 4 being

The Benchmark does not mean Plateau the

problem with this and like I said before

this is a business which mean things are

going to change GPT 4 is currently the

Benchmark which means that companies are

incentivized to train their models to

surpass GPT 4 and then release that

model the reason this creates the

illusion that things are plateauing

around GPT 4 is because these companies

are no longer incentivized to go duly

pass GPT 4 they're only incentivized to

just beat it and that is because of

course with GPT 4 that is something that

people State oh it's the best system is

the best system so if another company

like gemini or anthropic or Google can

come out and say look our system

surpasses gbt 4 or benchmarks they're

going to immediately release that model

after it's fine- tuned or after after

it's whatever they've done with it and

then run with that so that they can

Market that and get the customer base

because they know that open AI are

waiting to release GPT 5 potentially

after the elections and that gives them

some time to reclaim the market share

understand that where GPT 4 is is just

an indication of where other models are

going to stop and if you think that that

is just a pure speculative argument look

at how close gp4 is to some of these

models you have to understand that if

they didn't beat gp4 they wouldn't be

releasing these models this is

86.4% they literally got it up to 86.8%

another one here 92% they got this up to

95% okay it's not like it's completely

surpassing them and I guess some people

like look it all you know slows down

around here no they just want it to be

as close as possible so that they can

get this out as quickly as possible

because they know by the time open hour

releases next again they're going to be

even behind and you can see here that

Gemini Ultra a lot of people were even

debating this because this was Chain of

Thought at 32 um and that what they did

in order to beat this because I'm

guessing that when they had finished

training the model and when they

finished fine-tuning it they had to you

know increasingly developed certain

methods just to get this metric right

here and that's why I state that gp4

being the Benchmark does not mean we're

currently at a plateau at all because

it's likely that these companies are

just benchmarking their models up to GPT

4 so that they can get them out now

here's why things are going to go even

crazy and remember I said this because

the next 365 days are going to be

absolutely insane agents are still early

okay and someone actually recently

created a benchmark where they're

talking about multimodal agents for

open-ended tasks in real computer

environments and essentially with this

you can see that humans can accomplish

over 72% of the task and the best

current AI agent can only do

12.24% what happens when AI agents can

get to above 80% that is truly going to

change everything and with the advanced

reasoning and with the advanced

capabilities of future models we going

to see a future that we've never seen

before now there was also something that

I covered in a previous video that I'm

guessing the majority of people are just

completely glossing over and I'm sure

it's because during the video I was kind

of sick because I had some kind of flu

whatever but I still made the video

anyways so essentially there was this

thing right here okay this is mesa's kpu

now this is a little bit speculative

because they haven't released too much

information but if you check the

benchmarks here you can see that this

surpasses claw 3 Opus Gemini Ultra and

Mr large at all benchmarks okay and this

is because they use a advanced reasoning

engine on top of gp4 Turbo now this is

pretty interesting because this shows us

that we are still very early on the

reasoning capabilities which is why I

argue that samman here says that gp4 is

dumb and why he also says here that it

was not a very good AI system so there

was one demo released by M's kpu in

which they showcased an AI system

actually doing reasoning with an

advanced task and recently on their

Twitter I'm not sure why it's not

getting any love or any actual you know

tweets about it they've shown that this

system is able to you know use some

reasoning steps and this is their system

they're messing around with it and

they're showing that it's able to

complete a lot of tasks really really

correctly so you have to remember that

internally things are going at light

speed things like qar things like other

companies now trying to get a piece of

the pie uh incorporating different

reasoning engines on top of gbt 4 are

going to push things further and

remember it was recently that Andrew NG

actually spoke about agentic workflows

and basically said that GPT 3.5 zero

shot was

48.3 five and of course it was Andrew NG

that did some research and found out

that GPT 3.5 zero shot was 48.1% correct

GPT 4 zero shot does better at 67% but

the Improvement was dwarfed by

incorporating an iterative agent

workflow and wrapped in an agent Loop

GPT 3.5 gets up to 95.1% so the point

here is that there's still a lot of

different architectures that we haven't

fully explored with some of the AI

systems that we do have which means that

we are far far far away from any sort of

plateau and things are going to keep

increasing number one we've got the gpus

increasing in terms of efficiency we've

got the data centers we've got all of

these things getting increasingly better

and of course we've got the fact that

internally open AI they are blisteringly

so far ahead that I'm guessing that

things are going to be shocking when

they are finally released