Is programming dead? | Stephen Wolfram and Lex Fridman

Summary

The transcript explores the evolving relationship between humans and computation, suggesting a future where users interact with AI without needing to understand underlying code. It discusses the historical shift in computer science education and speculates on the skills needed as AI becomes more integrated into daily tasks. The conversation touches on the potential for 'prompt engineering' and the use of psychological techniques to interact effectively with AI, hinting at a future where understanding AI might require new forms of 'mind hacks'.

Takeaways

• 🎨 The average art history student may not be familiar with programming but could benefit from learning computational thinking as it becomes more integrated into various fields.
• 🔍 There's a potential future where people interact with technology without needing to understand the underlying code, trusting the system to generate correct outputs.
• 💡 The ability to generate and test code through natural language prompts could lead to a shift in how we approach learning and using computational tools.
• 📚 The historical development of computer science education has seen a shift from theoretical to practical, driven by the demand for programming skills in various fields.
• 🌐 The rise of 'computational X' in different disciplines has led to an increased need for understanding computation, even if the specific programming languages change.
• 🚗 The analogy of driving a car without needing to know the mechanics suggests that understanding the outcomes and possibilities of computation might be more important than the technical details.
• 📝 Good prompt engineering, similar to good expository writing, is crucial for effectively communicating with AI and getting the desired results.
• 🤔 The conversation raises questions about the future of computer science education and what skills will be necessary as AI and automation become more prevalent.
• 🧠 The discussion suggests that understanding the 'science' behind AI models could lead to new and unexpected ways of interacting with and manipulating them.
• 🔑 The concept of 'jailbreaking' AI, or finding unconventional ways to interact with AI systems, is an interesting area for exploration as we learn more about how these systems work.

Q & A

What is the speaker's view on how computational tools will evolve in the future?

The speaker believes that computational tools will become so user-friendly that people will interact with them using natural language, without needing to understand the underlying programming. Eventually, users will rely on the results or tests generated by the system rather than needing to look at or understand the code itself.

How does the speaker compare learning computational tools to learning mathematics?

The speaker notes that unlike mathematics, where you must learn the subject before applying it, computational tools might allow people to use them effectively before fully understanding how they work. This reverses the traditional learning process.

What does the speaker say about the historical development of computer science education?

The speaker reflects on how computer science departments were initially small and theoretical, focusing on topics like automata theory and compiler theory. Over time, they evolved to include more practical software engineering and IT-related skills as various fields became more computational.

Why does the speaker question the need to learn programming languages in the future?

The speaker suggests that if computational tools become sophisticated enough to handle tasks without requiring users to understand programming language specifics, then learning the mechanics of programming might become unnecessary. Instead, people will focus on higher-level goals and conceptualizing what they want to achieve.

What does the speaker mean by saying people need to know where they want to 'drive the car'?

The speaker uses the metaphor of driving a car to explain that people won’t need to understand the detailed mechanics of computation, but they will need to know what they want to achieve—i.e., have a clear sense of the objectives and possibilities within computational frameworks.

What connection does the speaker make between expository writing and prompt engineering?

The speaker argues that good expository writers, who excel at explaining concepts clearly, may also become effective prompt engineers. Crafting clear and precise prompts for AI systems requires similar skills to explaining things well in writing.

How does the speaker reflect on the fate of certain academic disciplines like geography or linguistics?

The speaker notes that some fields, like geography and linguistics, have disappeared or diminished in many universities just as their practical applications, such as GIS for geography, became more common. This observation suggests that the relevance of academic fields can fluctuate based on external trends.

Why does the speaker believe prompt clarity is important when interacting with AI?

The speaker notes that being sloppy when writing prompts for AI can lead to confusion or incorrect outputs. They emphasize the importance of clarity, as AI systems often mirror how humans respond to clear versus unclear instructions.

How does the speaker draw parallels between AI prompt hacking and human psychological techniques?

The speaker compares prompt hacking to psychological methods, such as those employed by therapists or in human interactions. Techniques like game-theoretic thinking or thought experiments, which challenge people mentally, can also help elicit better responses from AI systems.

What is the speaker curious about regarding the science of large language models (LLMs)?

The speaker is curious if there will be 'mind hacks' for LLMs, similar to optical illusions or psychological tricks for humans. They wonder if there are undiscovered methods of manipulating LLMs that go beyond human-like strategies, based on understanding the inner workings of the models.

Outlines

00:00

💡 The Evolution of Programming and Learning

The speaker discusses the changing landscape of programming and computer science education. They note that traditionally, art history students might not have been exposed to programming, but with advancements, people can now interact with computational systems intuitively without extensive prior knowledge. The speaker suggests that as AI and automation improve, people might trust the output of computational systems without needing to understand the underlying code. They also reflect on the historical development of computer science departments, which have shifted from theoretical to more practical and widely applicable skills. The speaker questions what people should learn in a world where computation is more accessible, pondering whether the focus should be on understanding the broader possibilities of computation rather than the technical details.

05:01

🚗 The Role of Prompt Engineering in AI Interaction

This paragraph delves into the concept of 'prompt engineering,' which is the art of effectively communicating with AI systems using natural language. The speaker hypothesizes that good expository writers might excel at crafting prompts that guide AI systems to produce desired outcomes. They compare the process to psychotherapy, where certain manipulative or thought-provoking techniques can elicit responses from the AI. The speaker is intrigued by the parallels between human communication and AI interaction, suggesting that the same expository skills that work with humans are also effective with AI. They also speculate about the potential for 'jailbreaking' AI, using unconventional methods to extract or influence information from AI systems, and ponder the future development of AI 'hacks' that could be as simple or as complex as psychological tricks.

10:03

🎓 The Transformation of Computer Science Education

The speaker reflects on their personal experience with the evolution of computer science departments in universities. They note the historical context, where computer science was once a niche field taught in small departments, focusing on theoretical aspects like automata theory and compiler theory. However, with the rise of computational applications across various fields, there was a surge in demand for practical programming skills, leading to a significant shift in computer science education. The speaker contemplates the future of computer science as a discipline, especially with the advent of AI and natural language interfaces, which may redefine what it means to be proficient in computation and how it is taught.

Keywords

💡Computational Language

Computational language refers to the set of formal languages used in computer programming. It is the medium through which programmers instruct computers to perform specific tasks. In the video, the speaker suggests that as AI becomes more advanced, people might interact with computational languages without needing to understand them deeply, similar to how they use apps without knowing the underlying code.

💡Art History Student

An art history student is someone studying the history, development, and interpretation of visual arts. The script uses this term to illustrate a contrast between traditional academic disciplines and the emerging need for computational skills. It suggests that art history students, who typically wouldn't be expected to know about programming, might find themselves needing to understand computational processes in the future.

💡Natural Language Processing (NLP)

Natural Language Processing is a field of AI that focuses on the interaction between computers and human language. It aims to enable computers to understand, interpret, and generate human language in a valuable way. The video discusses how NLP could allow users to interact with computational processes using natural language, making it more accessible.

💡Prompt Engineering

Prompt engineering is the art of crafting input prompts that effectively guide AI systems to produce desired outputs. The video suggests that as AI systems become more sophisticated, the skill of writing clear and effective prompts will become increasingly important, much like the skill of expository writing.

💡Expository Writing

Expository writing is a mode of writing that explains, informs, or describes a subject. It is often used in academic and professional settings to convey complex ideas clearly. The video posits that good expository writers may make good 'prompt engineers' because they can articulate instructions and questions in a way that AI systems can understand.

💡AI Wranglers

AI Wranglers is a term used in the video to describe individuals who specialize in managing and directing AI systems. It implies a role that is part programmer, part psychologist, understanding both the technical aspects of AI and the nuances of human interaction.

💡Jailbreaking

In the context of the video, 'jailbreaking' refers to the practice of finding and exploiting vulnerabilities in AI systems to achieve outcomes that might not be intended by the system's designers. It's a metaphor for creative problem-solving and pushing the boundaries of what AI can do.

💡Phishing

Phishing is a type of cyber fraud where attackers attempt to acquire sensitive information by disguising themselves as a trustworthy entity. The video draws a parallel between phishing attacks on humans and potential 'fishing' of AI systems, suggesting that as AI becomes more integrated into our lives, new forms of manipulation and security threats may emerge.

💡Reverse Engineering

Reverse engineering is the process of deconstructing a system to understand its workings, often with the goal of replicating or improving upon its design. In the video, reverse engineering is discussed in the context of understanding AI systems, particularly in terms of figuring out how to communicate effectively with them.

💡AI Psychologists

AI psychologists is a term coined in the video to describe a potential future profession focused on understanding the 'psychology' of AI systems—how they process information, learn, and make decisions. It suggests a field that would blend psychology with AI to optimize human-AI interaction.

💡Computational X

Computational X is a term used in the video to describe the trend of applying computational methods to various fields (e.g., computational biology, computational finance). It reflects the growing importance of computational skills across academia and industry, and the potential for AI to transform these fields.

Highlights

The average art history student may not be familiar with programming, but advancements could allow for intuitive use without prior knowledge.

In the future, users might interact with computational systems through visual comparison and generation of code without understanding the underlying language.

There's a possibility that people will trust computational language code generated by AI without needing to understand it.

AI could generate tests, and users might only need to verify the results rather than understand the code itself.

The role of understanding computational language may diminish as AI improves, with users focusing more on the outcome.

The historical development of computer science education has shifted from theoretical to practical, reflecting changes in industry needs.

The demand for IT and software engineering skills in the 90s led to a pivot in computer science education.

The idea emerged that learning programming is essential to engage in computational fields, which may not be the case in the future.

The future may see a shift from learning the mechanics of programming to understanding what is computationally possible.

The importance of expository writing skills in creating effective prompts for AI systems.

The comparison between human interaction and AI interaction, suggesting AI can be manipulated using similar psychological tactics.

The potential for 'jailbreaking' AI systems through novel techniques that exploit the AI's understanding of natural language.

The possibility that understanding the science behind AI could lead to new, unconventional methods for interacting with AI systems.

The concept of 'AI Wranglers' or 'AI psychologists' may emerge as new professions focused on optimizing AI interactions.

The current 'prompt hacks' are human psychological in nature, but future hacks might be more technical and specific to AI systems.

The accessibility of reverse engineering AI through natural language interfaces, allowing a broader population to participate.

The speaker's unique perspective on witnessing the rise and potential transformation of computer science departments in universities.

Transcripts

right now you know the average you know

art history student or something

probably isn't going to you know they're

not probably they don't think they know

about programming and things like this

but by the time it really becomes a kind

of purely you know you just walk up to

it there's no documentation you start

just typing you know compare these

pictures with these pictures and you

know see the use of this color whatever

and you generate this piece of of

computational language code that gets

run you see the results you say oh that

looks roughly right or you say that's

crazy

um and maybe then you eventually get to

say well I better actually try and

understand what this computational

language code did

um and and that becomes the thing that

you learn just like it's kind of an

interesting thing because unlike with

mathematics where you kind of have to

learn it before you can use it this is a

case where you can use it before you

have to learn it well I got a sad

possibility here or maybe exciting

possibility that very quickly people

won't even look at the a computational

language they'll trust that it's

generated correctly as you get better

and better generating that language uh

yes I think that there will be enough

cases where people see you know because

you can make it generate tests too yes

and and so you'll say

um we're doing that I mean that's it's a

pretty cool thing actually yes because

you you know say this is the code and

you know here are a bunch of examples of

running the code yeah okay people will

at least look at those and they'll say

that example is wrong and you know then

it'll kind of wind back from there and I

agree that that the the kind of the

intermediate level of people reading the

computational language code in some case

people will do that in other case people

just look at the tests

and or even just look at the results and

sometimes it'll be obvious that you got

the thing you wanted to get because you

were just describing you know make me

this interface that has two sliders here

and you can see it has that those two

sliders there and that's that's kind of

that's that's the result you want but I

I think you know one of the questions

then is in that setting where you know

you have this kind of ability broad

ability of people to access computation

what should people learn you know in

other words right now you you know you

go to Computer Science school so to

speak and a large part of what people

end up learning I mean it's been a funny

historical development because back you

know 30 40 years ago computer science

departments were quite small and they

taught you know things like final

automata Theory and compiler Theory and

things like this

um you know company like mine rarely

hired people who'd come out of those

programs because the stuff they knew was

I think it's very interesting I love

that theoretical stuff but um you know

it wasn't that useful for the things we

actually had to build build in software

engineering and then kind of there was

this big pivot in the in the 90s I guess

where there was a big demand for sort of

I.T type programming and so on and

software engineering and then you know

big demand from students and so on you

know we want to learn this stuff and uh

and and I think you know the thing that

really was happening in part was lots of

different fields of human endeavor were

becoming computational you know for all

acts there was a there was a

computational x and this is a um uh and

that was the thing that um that people

were responding to

um and but then kind of this idea

emerged that to get to that point the

main thing you had to do was to learn

this kind of trade or or skill of doing

you know programming language type

programming and and that uh you know it

kind of is a strange thing actually

because I you know I remember back when

I used to be in the professor in

business which is now 35 years ago so

gosh that's rather long time

yeah

um you know it was it was right when

they were just starting to emerge kind

of computer science departments that

sort of a fancy research universities

and so on I mean some had already had it

but the the other ones that that um were

just starting to have that and it was

kind of a a thing where they were kind

of wondering are we going to put this

thing that is essentially a a trade-like

skill are we going to somehow attach

this to the rest of what we're doing and

a lot of these kind of knowledge work

type activities have always seemed like

things where that's where the humans

have to go to school and learn all this

stuff and that's never going to be

automated yeah and you know this is It's

kind of shocking that rather quickly you

know a lot of that stuff is clearly

automatable and I think you know but the

question then is okay so if it isn't

worth learning kind of uh you know how

to do car mechanics you only need to

know how to drive the car so to speak

what do you need to learn and you know

in other words if you don't need to know

the mechanics of how to tell the

computer in detail you know make this

Loop you know set this variable but you

know set up this array whatever else if

you don't have to learn that stuff you

don't have to learn the kind of under

the hood things what do you have to

learn I think the answer is you need to

have an idea where you want to drive the

car in other words you need to have some

notion of you know your you know you

need to have some picture of sort of

what the what the architecture of what

is computationally possible is well

there's also this kind of artistic

element of um of conversation because

you ultimately you use natural language

to control the car

so it's not just the where you want to

go well yeah you know it's interesting

it's a question of who's going to be a

great prompt engineer yeah okay so my

current theory this week good expository

writers are good prompt Engineers what's

an expository range so like uh somebody

who can explain stuff well huh police

department does that come from in the

University yeah I have no idea I think

they killed off all the expository

writing departments well there you go

strong Wars receivable from well I don't

know I don't I'm not sure if that's

right I mean I I actually am curious

because in fact I just sort of initiated

this kind of study of of what's happened

to different fields at universities

because like you know there used to be

geography departments at all

universities and then they disappeared

actually right before GIS became common

I think they disappeared you know

Linguistics departments came and went in

many universities it's kind of

interesting because these things that

people have thought were worth learning

at one time and then they kind of die

off and then you know I do think that

it's kind of interesting that for me

writing prompts for example well I

realize you know I think I'm an okay

expository writer and I realize when I'm

sloppy writing a prompt and I don't

really think because I'm thinking that

I'm just talking to an AI I don't need

to you know try and be clear in

explaining things that's when it gets

totally confused and I mean in some

sense you have been writing prompts for

a long time with wolf from alpha

thinking about this kind of stuff yeah

how do you convert natural language into

competition well right but that's a you

know the one thing that I'm wondering

about is uh you know it is remarkable

the extent to which you can address an

llm like you can address a human so to

speak and and I think that is because it

you know it learned from all of us

humans it's it's uh the reason that it

responds to the ways that we will

explain things to humans is because it

is a representation of how humans talk

about things but it is bizarre to me

some of the things that kind of are sort

of expository mechanisms that I've

learned in trying to write clear you

know expositions in English that you

know just for humans that those same

mechanisms seem to also be useful for

for for the llm but on top of that

what's useful is the kind of mechanisms

that maybe a psychotherapist employs

which is a kind of uh like almost

manipulative or game theoretic

interaction where Maybe

you would do with a friend like a

thought experiment that if this was the

last day you were to live or yeah if if

I ask you this question and you answer

wrong I will kill you those kinds of

problems seem to also help yes in

interesting ways yeah so it makes you

wonder like the way a therapist I think

would like a good therapist probably you

we create layers

in our human mind to between like uh

between between the outside world and

we'll just true what is true to us and

um maybe about trauma and all those

kinds of things so projecting that into

an llm maybe there might be a deep truth

that's it's concealing from you it's not

aware of it you get to that truth you

have to kind of really kind of

manipulate this yeah yeah right it's

like these jailbreaking jailbreaking for

llms and but the space of jailbreaking

techniques

as opposed to being fun little hacks

that could be an entire system sure yeah

I mean just think about the computer

security aspects of of how you you know

phishing and and computer security you

know fishing of humans yeah and fishing

of llms is is a is a they're very

similar kinds of things but I think I

mean this this um

uh you know this whole thing about kind

of the AI Wranglers AI psychologists all

that stuff will come the thing that I'm

curious about is right now the things

that are sort of prompt hacks are quite

human they're quite sort of

psychological human kinds of hacks the

thing I do wonder about is if we

understood more about kind of uh the

science of the llm will there be some

totally bizarre hack that is you know

like repeater word three times and put a

this that and the other there that

somehow plugs into some aspect of how

the llm works

um that is not you know that that's kind

of like like an optical illusion for

humans for example like one of these

mind hacks for humans what are the Mind

hacks for the llms I don't think we know

that yet and that becomes a kind of

us figuring out reverse engineering the

language that controls the L alums and

the thing is the reverse engineering can

be done by a very large percentage of

the population now because it's natural

language interface right it's kind of

interesting to see that you were there

at the birth of the computer science

department

as a thing and you might be there at the

death of the computer science term as

the thing well yeah I don't know there

were computer science departments that

existed earlier but the ones the

broadening of every University had to

have a computer science department yes I

was I was uh I watched that so to speak