Loading video...

Video Failed to Load

There was a problem loading this video. This could be due to a temporary network issue or the video might be unavailable.

We’re bringing Aristotle to code. AI is writing more code than anyone can review. We're building the AI that verifies it. Starting today with C and Verilog.

Harmonic

27,625 subscribers

211,896 views • 5 days ago •via X (Twitter)

Science & Technology

Anya Rossi• Live Now

Private livecam show

21 Comments

2147M5 days ago

.@HarmonicHQ actually did this too!

Harmonic5 days ago

Same rule as math: no partial credit. You tell Aristotle what your code is supposed to do, in plain English. It either proves it, or shows you where it doesn't.

Harmonic5 days ago

You don't need to be a formal verification expert to use it. What used to take years to master can now be run by any engineer on every commit.

Harmonic5 days ago

C and Verilog are just the start. More languages are on the way. Try Aristotle today:

Bayesian5 days ago

you can write your own script, AI is not actually good at this yet imo

manofmen (LATE)5 days ago

🔥🔥🔥🔥🔥🔥 @HarmonicAgents gonna run this Great work @tachim and team

Aceinyahole🐦‍🔥5 days ago

@HarmonicHQ has brought Aristotle onchain!

Tyler Kenney ♔5 days ago

I do love math

WAO5 days ago

Fireee

LoreWeaver5 days ago

Ok, Verilog on day one is a flex. Half of what my agents write runs fine and still feels off, and I'd happily pay for something that catches that before I do.

Tally | x402 data & agent credit5 days ago

Good tech

Michael Machuca5 days ago

nice

Mehmet Mars Seven 🐴5 days ago

Congrats, and thanks for the free tokens!

Arnav ✈️ EMNLP 20265 days ago

Is the input by the user code or a specification? Is the interface a user converges with a formal system on the specification and then a “prover” generates a program that satisfies this specification?

Peng Cheng, asking λP2-λC5 days ago

Congrats! long overdue. meanwhile can we also have the Lean 4.35 upgrade? Old version is already like a sieve

Marco Pinnisi5 days ago

It's a shitty finetune, isn't it? I bet it doesn't even have Bend 2 training data which is what people care about now

Khurram Saleem5 days ago

sick

Aapakari5 days ago

Verilog first, spicy. Let's see it handle a race condition.

Tomas Gonzalez5 days ago

🪐

Conpertus Tools5 days ago

@dhh thoughts and guidance?

T3ch5 days ago

hey, @tachim why is 5 credits which supposedly cost 500 dollars drain so fast? it couldnt last 2 hours of work for me!

Related Videos

Today we're launching REVIEW.FAST, speeding up human review of AI generated code 3X. AI changed the world, sped up writing code, but shifted the bottleneck to review. If you can speed up code review, you can ship more. It's as simple as that.

Today we're launching REVIEW.FAST, speeding up human review of AI generated code 3X. AI changed the world, sped up writing code, but shifted the bottleneck to review. If you can speed up code review, you can ship more. It's as simple as that.

Louis Knight-Webb

46,349 views • 10 months ago

AI is changing the software engineering craft. Anders Hejlsberg (Anders Hejlsberg) - creator of C#, TypeScript and industry legend - on why code review needs to get more enjoyable in response: #1 - AI is shifting the craft from writing code, to reviewing code: "In a sense, we're all turning into project managers. We can have an army of junior programmers, called agents, that will just spit out reams of code but someone's got to have the big picture and review all of that. And so, increasingly, our craft is going from one of writing the code, to one of reviewing the code and building the architecture of the code and overseeing the work. It's a different kind of craft. It's a different kind of enjoyment. I've always liked writing the code. To me that was the fulfilling part, seeing it work. In a way, AI robs a little bit of that, because I am less interested in reviewing code." #2 - The code review experience should be improved: "I think we could also make the process of reviewing code much more interesting than it is today. I mean, today, you see a list of diffs in alphabetical order and now it's up to you to make heads or tails of it. There are more pedagogical ways of presenting that. And you could have commentary generated by the AI that tells you what the changes are and whatever, and then tries to guide you along. So that symbiotic relationship, I think we need to work on that more and to keep the enjoyment in there."

AI is changing the software engineering craft. Anders Hejlsberg (Anders Hejlsberg) - creator of C#, TypeScript and industry legend - on why code review needs to get more enjoyable in response: #1 - AI is shifting the craft from writing code, to reviewing code: "In a sense, we're all turning into project managers. We can have an army of junior programmers, called agents, that will just spit out reams of code but someone's got to have the big picture and review all of that. And so, increasingly, our craft is going from one of writing the code, to one of reviewing the code and building the architecture of the code and overseeing the work. It's a different kind of craft. It's a different kind of enjoyment. I've always liked writing the code. To me that was the fulfilling part, seeing it work. In a way, AI robs a little bit of that, because I am less interested in reviewing code." #2 - The code review experience should be improved: "I think we could also make the process of reviewing code much more interesting than it is today. I mean, today, you see a list of diffs in alphabetical order and now it's up to you to make heads or tails of it. There are more pedagogical ways of presenting that. And you could have commentary generated by the AI that tells you what the changes are and whatever, and then tries to guide you along. So that symbiotic relationship, I think we need to work on that more and to keep the enjoyment in there."

The Pragmatic Engineer

39,073 views • 5 months ago

Let’s make an AI photo generator app with Cursor in 7min. Just give Cursor Composer a prompt file and the AI will write 100% of the code. AI is making it so that *anyone* can start building software today. Full prompt below.

Let’s make an AI photo generator app with Cursor in 7min. Just give Cursor Composer a prompt file and the AI will write 100% of the code. AI is making it so that anyone can start building software today. Full prompt below.

Mckay Wrigley

426,837 views • 2 years ago

is a new superhuman full stack engineer that 2 weeks after launch has users building 4000 new products thanks to AI each day. It let's anyone build 20x faster than writing code. And it's free to try.

is a new superhuman full stack engineer that 2 weeks after launch has users building 4000 new products thanks to AI each day. It let's anyone build 20x faster than writing code. And it's free to try.

Anton Osika

264,525 views • 1 year ago

👋 Meet Duet AI for Developers! Start writing code at the speed of thought with #AI code completion, code generation, and a chatbot where you can ask any code-related questions in natural language. ✨ Try it today at no cost →

👋 Meet Duet AI for Developers! Start writing code at the speed of thought with #AI code completion, code generation, and a chatbot where you can ask any code-related questions in natural language. ✨ Try it today at no cost →

Google Cloud Tech

19,156 views • 2 years ago

VS Code: the open source AI editor. Starting today.

VS Code: the open source AI editor. Starting today.

Visual Studio Code

453,737 views • 1 year ago

Nobody is writing 90% of their code using AI. Here's the uncomfortable truth: The real productivity gain from using AI to write code is closer to 10%, nowhere near the 90% people claim. Sundar Pichai said in 2024 that 30% of the new code at Google was AI-generated. However, he went on to admit, during Lex Friedman's podcast, that engineering velocity had only increased by about 10%. AI-generated code isn't free code. It still has to be reviewed, tested, and made production-ready. Optimizing a single step (code generation) doesn't boost output if bottlenecks shift elsewhere (code reviews). It doesn't matter how much code you generate if you can't keep up the review process. The solution: Automate as much as you can the review and verification of your code. I'm working with Sonar, who is sponsoring this post, and they will take care of the code quality and security analysis of your code: • They review over 300B lines of code every single day • They cover reliability, security, and maintainability for your code • You can integrate them into your CI/CD pipeline • You can install them in your IDE (I use their VSCode extension) • Support for more than 30 languages Here is a link so you can check them out:

Nobody is writing 90% of their code using AI. Here's the uncomfortable truth: The real productivity gain from using AI to write code is closer to 10%, nowhere near the 90% people claim. Sundar Pichai said in 2024 that 30% of the new code at Google was AI-generated. However, he went on to admit, during Lex Friedman's podcast, that engineering velocity had only increased by about 10%. AI-generated code isn't free code. It still has to be reviewed, tested, and made production-ready. Optimizing a single step (code generation) doesn't boost output if bottlenecks shift elsewhere (code reviews). It doesn't matter how much code you generate if you can't keep up the review process. The solution: Automate as much as you can the review and verification of your code. I'm working with Sonar, who is sponsoring this post, and they will take care of the code quality and security analysis of your code: • They review over 300B lines of code every single day • They cover reliability, security, and maintainability for your code • You can integrate them into your CI/CD pipeline • You can install them in your IDE (I use their VSCode extension) • Support for more than 30 languages Here is a link so you can check them out:

Santiago

296,928 views • 1 year ago

The best way to learn about AI is to build an AI app yourself. With tools like Bubble and OpenAI, you can create apps like an AI chatbot without writing code 🤯 Stay ahead on AI and start learning today:

The best way to learn about AI is to build an AI app yourself. With tools like Bubble and OpenAI, you can create apps like an AI chatbot without writing code 🤯 Stay ahead on AI and start learning today:

Seth Kramer

2,614,004 views • 3 years ago

After reaching AI-native terminal velocity we are swamped in code review. The core insight is that as long as you can write code in parallel you will continue to be stuck until you can review it in parallel. I was surprised by how quickly this became my daily driver for not just review code, but writing it as well.

After reaching AI-native terminal velocity we are swamped in code review. The core insight is that as long as you can write code in parallel you will continue to be stuck until you can review it in parallel. I was surprised by how quickly this became my daily driver for not just review code, but writing it as well.

Evan Boyle

23,986 views • 8 months ago

LeetCode is dead. Developers don't write code line-by-line anymore. They orchestrate AI agents working in parallel, review AI-generated code, and make architectural decisions. That's the job now. But most interview processes haven't caught up. They still test algorithm memorization instead of AI fluency, code review, and judgment. We're building assessments for next-gen hiring that mirror how developers actually work. Here's how we think about it:

LeetCode is dead. Developers don't write code line-by-line anymore. They orchestrate AI agents working in parallel, review AI-generated code, and make architectural decisions. That's the job now. But most interview processes haven't caught up. They still test algorithm memorization instead of AI fluency, code review, and judgment. We're building assessments for next-gen hiring that mirror how developers actually work. Here's how we think about it:

HackerRank

344,947 views • 6 months ago

Why code review is the single most important category in software “I actually think code review is the single most important category today in software. By far the most important category. The rate at which we are building code is unprecedented, but it is never production-ready or production-intent. The amount of time it takes to review, process it and make it production-intent is still an order of magnitude more than building. It is much more important than code generation.” Jason ✨👾SaaStr.Ai✨ Lemkin Love to hear your thoughts Daksh Gupta Harjot Gill Itamar Friedman Zach Lloyd

Why code review is the single most important category in software “I actually think code review is the single most important category today in software. By far the most important category. The rate at which we are building code is unprecedented, but it is never production-ready or production-intent. The amount of time it takes to review, process it and make it production-intent is still an order of magnitude more than building. It is much more important than code generation.” Jason ✨👾SaaStr.Ai✨ Lemkin Love to hear your thoughts Daksh Gupta Harjot Gill Itamar Friedman Zach Lloyd

Harry Stebbings

23,053 views • 9 days ago

AI agents are writing more code. How can developers ensure that code is correct? Microsoft Research’s Proof-Driven Development (PDD) Kit explores a more rigorous approach: it translates a developer’s intent into a formal specification, then uses formal verification to prove that generated code meets it. PDD Kit works with Verus for Rust and is integrated with GitHub Next’s Chopin, bringing formal methods into an AI-assisted development workflow for teams. See the work in action at GitHub Universe, Oct. 28–29, 2026.

AI agents are writing more code. How can developers ensure that code is correct? Microsoft Research’s Proof-Driven Development (PDD) Kit explores a more rigorous approach: it translates a developer’s intent into a formal specification, then uses formal verification to prove that generated code meets it. PDD Kit works with Verus for Rust and is integrated with GitHub Next’s Chopin, bringing formal methods into an AI-assisted development workflow for teams. See the work in action at GitHub Universe, Oct. 28–29, 2026.

Microsoft Research

67,015 views • 11 days ago

"The one thing that's going to be truly future proof is judgment. What is the biggest challenge when you have a thousand AI engineers writing code? AI slop. In an era when you can do everything, the question is which of these things matter. On the engineer side, it is evaluating the code. Which of this code is even valuable? You can have AI engineers writing beautiful code that could be wrong, that could have bugs in it, that could be vulnerable. Someone needs to review it. Make sure to have a human review at some point, especially critical code that is in the core of your system. On the product side is judgment around what needs to be built and evaluating. And similarly in design, you have to have judgment around, does it make sense in the broader design system. I think this judgment is the number one thing that humans are going to bring in an era of infinite productivity."

"The one thing that's going to be truly future proof is judgment. What is the biggest challenge when you have a thousand AI engineers writing code? AI slop. In an era when you can do everything, the question is which of these things matter. On the engineer side, it is evaluating the code. Which of this code is even valuable? You can have AI engineers writing beautiful code that could be wrong, that could have bugs in it, that could be vulnerable. Someone needs to review it. Make sure to have a human review at some point, especially critical code that is in the core of your system. On the product side is judgment around what needs to be built and evaluating. And similarly in design, you have to have judgment around, does it make sense in the broader design system. I think this judgment is the number one thing that humans are going to bring in an era of infinite productivity."

Patrick OShaughnessy

24,818 views • 9 months ago

Code Four (Code Four) is building the AI Copilot for law enforcement officers. It auto-generates reports, verifies narratives, and surfaces evidence instantly, cutting desk time by 60% so officers can be where communities need them most. Congrats on the launch, george, Dylan Nguyen and @KnifeNate!

Code Four (Code Four) is building the AI Copilot for law enforcement officers. It auto-generates reports, verifies narratives, and surfaces evidence instantly, cutting desk time by 60% so officers can be where communities need them most. Congrats on the launch, george, Dylan Nguyen and @KnifeNate!

Y Combinator

22,167 views • 1 year ago

Writing formal specifications has always been the bottleneck in formal verification. Today we're launching AutoProver, an agentic formal verification system that reads your code, generates specifications, and verifies them. 🧵

Writing formal specifications has always been the bottleneck in formal verification. Today we're launching AutoProver, an agentic formal verification system that reads your code, generates specifications, and verifies them. 🧵

Certora

45,861 views • 3 months ago

Creator of C++, Bjarne Stroustrup: AI-generated code isn't ready — it generates more bugs, more bloat, more security holes, and is nearly impossible to validate "senior developers are already retiring rather than deal with it" The problem is that even a small prompt change can shift the entire codebase in unpredictable ways

Creator of C++, Bjarne Stroustrup: AI-generated code isn't ready — it generates more bugs, more bloat, more security holes, and is nearly impossible to validate "senior developers are already retiring rather than deal with it" The problem is that even a small prompt change can shift the entire codebase in unpredictable ways

Haider.

1,752,606 views • 5 months ago

“code is cheap, anyone can code with ai” sure, but the bottleneck has never been coding and it still never is. so if you misunderstand the real differentiating factor, you’re not gonna make it.

“code is cheap, anyone can code with ai” sure, but the bottleneck has never been coding and it still never is. so if you misunderstand the real differentiating factor, you’re not gonna make it.

ℏεsam

23,431 views • 9 months ago

Teams are cooking some code. 🧑‍💻 Smart Contracts, APIs, frontend, AI agents. It is possible to develop products on chain under 24 hours. Complete products. Especially with the age of AI, everything is faster, and anyone with some prompting skills and an idea can start building. The night will be long. Time to build more!

Teams are cooking some code. 🧑‍💻 Smart Contracts, APIs, frontend, AI agents. It is possible to develop products on chain under 24 hours. Complete products. Especially with the age of AI, everything is faster, and anyone with some prompting skills and an idea can start building. The night will be long. Time to build more!

Robert Sasu | dev/acc

17,582 views • 1 year ago

Cursor shipped a /thermo-nuclear-code-review for the TOUGHEST AI code review possible. But is it any good? Let's dive in:

Cursor shipped a /thermo-nuclear-code-review for the TOUGHEST AI code review possible. But is it any good? Let's dive in:

Matt Pocock

136,487 views • 5 months ago