text stringlengths 0 2.68k |
|---|
**Adam Stacoviak:** ...or Codex, even. They're always last on my list, for some reason... But you know, the moment you give them that slash command, now the agent has access to the CLI... If it's got great help docs, so --help on all your commands... Now the new frontier is like this really well fleshed out CLI... I fe... |
**Ajay Kulkarni:** No, it's a great point. And I think -- and again, hey, is MCP gonna be the end all protocol? I don't know. But there will be a protocol, and I think that's the point. That's the point you're making, and it's the point I'm trying to make. And why do you need this protocol? Well, number one, you need a... |
And then the CLI is another way to do that. The CLI is another way to give essentially a tool to the agent that is properly defined in what it can and can't do. Yeah, I think it's really interesting. I think the team's done some really cool -- I think this docs thing... And I'm not sure if I'm beating a dead horse here... |
I think that's been really cool, and I think that is -- it's almost a variant of the skills that you see in Claude, but I think that's another element of agents that I think is underexplored, is the fact that with the right markdown, they can become experts in anything. |
**Adam Stacoviak:** Did you see the talk from Barry Zhang and - I think his name's Mahesh Murag. They're both from Anthropic... It's the first time I've seen -- well, I guess the first time I thought I would see anybody talk on stage for an hour about markdown. They introduced -- |
**Ajay Kulkarni:** \[laughs\] I think I saw the tweet about it. It was like a packed room, or something... |
**Adam Stacoviak:** Yeah, an hour long talk about using markdown. You just wouldn't expect that. That's not the language I thought would be front and center. Now, I love reading markdown. I personally prefer markdown. I've been an Obsidian user for years now, and so I'm all about flat files, nothing that's sort of obfu... |
**Ajay Kulkarni:** \[01:14:03.02\] It is wild. In fact, I have a friend working on a markdown editor for vibe coding. Not for note taking, but for like vibe coding. No, it's a new world, man... Maybe we'll find -- that's the weird thing with these agents, is that in some ways they are very similar to us. You don't teac... |
**Adam Stacoviak:** Words... |
**Ajay Kulkarni:** Words. |
**Adam Stacoviak:** Words, man... Words. |
**Ajay Kulkarni:** Yeah, it is wild. |
**Adam Stacoviak:** I do feel that. |
**Ajay Kulkarni:** Well, I also think -- with your discussion earlier, I had a thought about the terminal, about how more people are using the terminal, living in the terminal... There's probably a business idea out there - I'm sure someone's working on it - of building the modern terminal. |
**Adam Stacoviak:** I think... Well, my favorite - and I'll give them a nod, because they are my favorite - is Warp. I think they've been... Gosh, I recall talking to them almost five years ago, and it was "The future of the terminal", I want to say... "The terminal of the future", I believe, was the title we shipped o... |
Now, I know Mitchell Hashimoto and Ghostty is awesome. I love it. I'm so thankful for what they're doing, and especially what Mitchell did with the foundation kind of piece to it, to kind of give Ghostty a home forever... But something about Warp just gets me, and so I just like Warp. So I think they may be ahead of th... |
**Ajay Kulkarni:** Yeah. Well, it sounds like Warp -- did Warp start before this whole AI thing? |
**Adam Stacoviak:** Oh, yeah. It's probably two years before, I want to say. At least two to three years before GPT-3 made people think that emails will forever be written by machines. And we're like "Nah, it'll never happen." You know what I mean? And now, "Can you summarize this into an email form for me? I will edit... |
\[01:17:14.00\] You can even teach your agents to think like you think. Especially in Claude, with these plugins and stuff like that, and the way you can change the prompt behavior to make it more antagonistic to your ideas, and not be like a yes agent. "Yes, Ajay, that was the most brilliant idea ever. Let's do this."... |
**Ajay Kulkarni:** It's a bicycle for your mind. With a mind. |
**Adam Stacoviak:** Yeah. And now we have it for our database. I think that's such a wild thing, man. But what's on the horizon for you? I know that we can't uncover things that are maybe releasing soon, or... I'm not asking you to spill the beans, but what's over the horizon? A little bit that's coming up soon, that y... |
**Ajay Kulkarni:** Yeah. Well, so we started off as Timescale, building a database that scaled Postgres for time-series workloads. And then it became a Postgres that scaled for workloads in general, especially analytical workloads. What we got into with agentic Postgres is now serving not just the scale use case, but t... |
We've actually been toying on a side project that is like a separate MCP server that is a little bit more than the database... And this is actually being developed by -- remember that engineer who said it would take a month and it took him 12 months? \[laughs\] It's the same guy doing this one. |
**Adam Stacoviak:** Wow. |
**Ajay Kulkarni:** But he's moving a lot faster. And we launched a new storage layer called Fluid Storage to enable kind of ephemeral forks... You'll see more stuff around that. Fast forking, safe forking, sandboxes... More around search, and more at scale. |
Yeah, and I think -- again, the job we do is "Hey, when you're a developer and you're looking for a database, we give you the database that just works." And it's built on Postgres, so you know you can trust it... But "just works" means something different at start than it does at scale. At start it means "Hey, it just ... |
**Adam Stacoviak:** That's a lot of people. I didn't expect that number. I should have asked you that sooner, but yeah, that's a lot of folks. Wow, that's a lot of folks behind you. Pretty wild. |
Well, Ajay, it's been a pleasure diving into where you've been, and why you've built what you've built, and how you've built what you've built. It's been fun. Thank you. |
**Ajay Kulkarni:** Thank you, Adam. Pleasure. Thanks for having me. Good to see you. |
• Salavatore Sanfilippo (Antirez) reflects on the early days of Redis and the evolution of the tech industry |
• He discusses the origins of Redis, starting from his work on security, embedded systems, and a Tcl interpreter |
• He talks about the early days of Web 2.0 in Italy and how he and a friend created services similar to Delicious and Digg |
• He describes how Telecom Italia approached them and asked to join their network, leading to a significant spike in traffic and the need for a caching solution |
• He explains how he created a memoization function in PHP and later developed Redis as a caching solution with a focus on persistence and low memory footprint |
• He credits his background in security and low-level programming with influencing the design of Redis and its persistence mechanism using the fork system call |
• Redis creator Salvatore Sanfilippo talks about the early days of Redis and its release on Hacker News |
• Ezra Morbius is mentioned as someone who helped Redis gain traction |
• Sanfilippo shares his approach to working on Redis, which involved focusing on subsystems that interested him and providing premium customer service to users |
• He discusses the "inflation" of Redis and the shift away from it due to the rise of strong consistency systems |
• Sanfilippo reflects on the importance of community involvement and treating users as customers, even if they are not paying |
• He shares his experience working with various companies, including VMware, Pivotal, and Redis Labs, and how they allowed him to work on Redis at his discretion |
• Sanfilippo emphasizes the importance of recognizing the value of community involvement and treating users as customers, even in open-source projects. |
• The creation and naming of Redis Labs and its relationship with Salvatore Sanfilippo |
• Salvatore Sanfilippo's departure as lead maintainer of Redis in 2020 |
• The licensing history of Redis, including the use of GPL, BSD, and SSPL |
• The impact of cloud computing on the business model of open-source software and the challenges of creating a business with a permissive license like BSD |
• The current discussion within Redis about the acceptance of the SSPL license and plans to make more features available in the community edition. |
• Redis' licensing updates and potential return to open source |
• Discussion of the AGPL license and its similarity to the SSPL license |
• Salvatore Sanfilippo's opinions on the company's licensing decision and his desire to switch back to open source |
• The impact of licensing changes on the community and Redis' relationship with its users |
• Plans for updating the license and potential cultural changes within the company |
• Salvatore's return to Redis and his role in promoting a more open and community-driven approach |
• The involvement of the CEO and other employees in supporting open source initiatives |
• The size and structure of the Redis company, with approximately 1,000 employees |
• Discussion of the size and global presence of Redis and its development team |
• Redis' development being almost completely done in Israel, with a strong pool of talented and loyal developers |
• Acquisition of new developers with exceptional skills, such as Oran and Yossi |
• Comparison with Valkey, including the design approach and AGPL licensing |
• Potential divergence of Redis and other projects, such as Valkey, and the benefits of competition in innovation |
• Discussion of Redis AI, specifically vector sets and learned embeddings |
• Mention of Claude Sonnet and its limitations, including the use of RAG and large prompts |
• Comparison with Open AI systems and their approach to handling large context and avoiding RAG |
• Frontier models are Mixture of Experts (MOE) architecture, which partitions the model into clusters and selects the relevant parts for each task |
• MOE models require more VRAM to store the entire model in RAM, but use less GPU |
• Large language models (LLMs) can be run on resource-constrained systems with sufficient VRAM, but are very slow |
• Tokenization and attention mechanisms can cause models to slow down with long context or complex input |
• Prompt engineering is becoming important again to optimize performance and efficiency with resource-constrained systems |
• A new Redis vector set is being developed to improve performance and efficiency, but no timeline is given for its release |
• Redis is planning to release vector embeddings in one to two months |
• Vector embeddings can be used for various tasks, including retrieving information from large datasets and performing math calculations |
• Use cases include tracking calories, generating text, and identifying similar items or objects |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.