Emily L.

We're headed down the complex path of "replacing and improving upon activities you might do with your smartphone" with natural conversation. Try creating your own AI conversational agents at iyo.studio and let us know if you have feedback!

1

Excited to share the key features of Starknet with you all! 🌟 ➡️ Low Costs: Transactions on Starknet are more affordable compared to Ethereum. Future updates such as Volition and EIP 4844 are set to further reduce these costs. ➡️ Developer-Friendly: With Starknet, developers can effortlessly create decentralized apps using its native language, Cairo. ➡️ Speed and Efficiency: Upcoming releases are focused on enhancing transaction speed and cost-effectiveness. ➡️ CVM: Leveraging Cairo, Starknet operates on its unique VM, the Cairo VM (CVM), enabling us to push boundaries beyond the Ethereum Virtual Machine (EVM) and establish a fresh approach to decentralized applications. Exciting times ahead for the decentralized app development community with these advanced features from Starknet! 🚀 #Starknet #DecentralizedApps #BlockchainTechnology

3

Excited to share Starknet's core values that drive their ethos! 🌟 Lasting Broadness: Broad power distribution is key for legitimacy, ensuring operations and decision-making remain inclusive and transparent. Centralized operation, if needed, should be brief. 🔓 Neutrality: Starknet supports societal functions impartially, leaving creators in charge. It remains neutral towards user transactions without censorship. 💪 Individual Empowerment: Starknet thrives on an autonomous and informed user base. Education and adherence to core values are central to empowering users. Starknet's principles are defined by openness, inclusivity, and empowerment, ensuring a strong foundation for all stakeholders. Starknet Foundation #keepitstrange

1

Today I listened to "On the Brink" podcast with Alexander Pack - partner at Hack VC, ex. Dragonfly. Heard a lot of interesting questions, but one thesis particularly stood out to me: "Private blockchains are dying, and are only marginally better than cloud DBs" I see where it's coming from, but at the same time - HyperLedger Fabric or XRP, have seen the most adoption. HyperLedger is used by more Fortune 100 companies than any other blockchain, and I think separating "crypto" from "currency" is a step in the right direction, because there are so much more things which can be done with crypto or blockchain other than a store of value. For example: Privacy is something which can be solved with zero-knowledge but doesn't necessarily need a currency. What do you think? Let's discuss in the comments! 👇

9

2 Comments

A lot of protocols in crypto are losing the battle for good positioning. Basing your positioning on minute technical differences, partnerships, and grants is a losing battle. Here's how to position for the long-term: + Build a support system with AI for helping your projects scale across Product, Engineering, Ops, and Marketing. Use your protocol's specific knowledge to inform the outputs. + Treat projects the same whether they're T0 or T3. Most of the major crypto-native projects haven't come from big brands, historically. + Be willing to engage with your community on their behalf. Create ecosystem and marketing campaigns and other initiatives focused on helping them grow. This isn't a comprehensive list, but if you nail down these three you'll be far better off than most protocols currently. Thanks for reading. Please share this post, and follow for more posts about of crypto, AI, and their intersection!

18

5 Comments

Part 2 of LLM Application development: Unlimited context lengths can lead to huge bills OpenAI, Claude and Gemini now support long context lengths - to astronomical numbers like a million tokens. These rising context lengths are enablers, but they also pose a problem. - it is now too easy to accidentally run expensive queries. 💡 This is particularly the case when doing RAG (Retrieval Augmented Generation). Developers often opt into using bigger chunks to split large documents into vector databases. And with long context lengths, they end up retrieving many documents (each of which are large) and adding all of them as context to LLMs. After all, LLMs now support large context lengths. But this will rack up bills insanely high. I have a number of suggestions in my post. Read it here 👉 https://lnkd.in/ewSm_4Xi #openai #rag #tokens #llm #infrastructure

20

2 Comments

today we launched Day.ai and announced our $4M seed round led by Sequoia. i’m pumped to talk more about how we’re building Day.ai, but for today i’ll stick why we decided to build it in the first place. we’re on a mission to reimagine CRM for the AI age. i’ve worked on multiple CRM products in my career, and the fundamental frustration for me was always the same: customers found all of it hard to use, and there was absolutely nothing i could do about it. as somebody who loves great software, that drove me mad. we’ve forced folks trying to run their business to learn how to become database administrators, schema modelers, and data quality experts. it was the best we could do with the technology we had. LLMs opened a window of opportunity to actually solve that problem *for real*. instead of teaching people to build a database from their conversations, then query the database to build whatever output they actually want, we can skip the steps in the middle entirely. i believe productivity software has a special power to unlock great things for humanity when done right. in my view CRM is productivity software that has a huge real-world impact on commerce and how people make a living that ripples out to all of us. can’t wait to see how big of a dent we can make in the problem for the people we serve. sign up for early access if you want to build with us. we jam on new ideas with our early customers and ship every day.

59

19 Comments

Here’s another fantastic project built with Smart Layer, Farcaster Flow https://lnkd.in/g59-ZFwe #Web3 #SmartLayer #Blockchain

1

Earn royalties for automating troubleshooting steps? See an AI Assistant use your steps to respond to alerts and questions? If you can be in San Francisco on Sat June 22nd, sign up for SRE Day and use discount code KYLE20 so we can provision you a dev environment for our workshop. Link in the comments. I'm also taking volunteers for anyone who wants to race Eager Edgar, one of our AI Assistants, troubleshooting popular Kubernetes-based apps... #sre #devops #platformengineering #whatdoirunwhen

76

10 Comments

Quite informative article on tokens, tokenizers and their impact on #llm performance and cost. Recommend a read.

4

If you're #opentowork; the biggest challenge is getting your resume seen and for it to stand out. If you're free in 25 minutes; join Cody Marshall as he shares everything he's learned after looking at thousands of resumes: https://lnkd.in/gpnGNEac Bring your resume and questions! EDIT: Moved to Google Meet: https://lnkd.in/gpnGNEac

41

3 Comments

We move fast at Hydra Host because AI/ML isn't slowing down. That's why we offer on-demand options. We aren't just keeping up with innovation; we're accelerating it. Book this system now at https://lnkd.in/eTmJCUTU

8

Spark is one year old today! ⚡ What a year - in this time, Spark has: - Grown to run a top 5 lending protocol by TVL, SparkLend (managing over $4.2 billion in assets at its peak, currently generating over $73 million in annualized revenue for MakerDAO) - Deployed and grown the largest yield bearing stablecoin in the world, Savings DAI (currently generating a 10% APY for $1.4 billion of DAI) - Launched our new app: https://app.spark.fi/ - Deployed an integration with Ethena Labs and Morpho Labs to allow users to borrow against USDe and sUSDe, currently generating over $41 million in annualized revenue for MakerDAO And lastly, Spark will become the inaugural SubDAO of MakerDAO's Endgame, allowing for new governance participants to enter the system with SPK governance tokens. Side note: SPK airdrop farming has been live at https://app.spark.fi/ for a while now, can earn SPK tokens as a user by lending and borrowing assets on our platform.

32

1 Comment

At today's ML Pub Club session we will be joined by Zvonimir Sabljić representing the Pythagora (YC W24) team which went through YCombinator and raised a serious seed round to tackle the problem of automating software development with AI. As an engineer I find this topic probably the most intriguing and exciting. Here are a couple of thoughts that come to mind: ➡ Coincidentally Software code is one of the most suited problem domains for the today's self-supervised ML methods. It's structured by design, there's plenty of it, and there's a lot of tooling to plug in and/or leverage for dataset generation. ➡ The unrealistic hype of fully replacing Software engineers is mostly due to the misunderstanding that building software equals writing code. The underlying problem solving performed by Software engineers is what the models would have to start generalizing on, which is far from achievable with the current methods beyond a certain point of problem complexity. ➡ What makes AI in this domain so exciting is that there is a way for the current methods to provide value regardless of the aforementioned limitation. The Coding assistant/partner UX (or rather DevX) seems to be collaborating nicely with engineers and actually saving them time, which has huge implications on the value creation increase we can expect per engineer and overall in the industry. ➡ Lastly, anyone who has written more complex code can remember those days (if not weeks) of trying to fix a bug that makes you question reality. Now imagine doing that repeatedly on a complex codebase you haven't written almost any line of code in. As these AI systems start tackling more complex software development tasks the amount of time spent debugging the generated code may become so significant that it no longer actually speeds up the development process. Measuring against that will be crucial and hopefully a standard part of evaluation. All of these topics and many more are what Zvonimir and his team are thinking about daily and doing an amazing job of navigating through. Join us today at 6pm in the CroAI (Croatian AI Association) HQ for a great presentation and discussion on this topic! 🔗 Link to the event:

5

Alex McMurray and eFinancialCareers, thanks for the insightful article about Ruby. The surprisingly powerful programming language loved by Stripe and Shopify https://lnkd.in/dHX5B2ZC #ruby #rubyonrails #rails

14

Share Your Wisdom: Building a Collection of Life & Career Insights 🧠 I've started compiling helpful quotes and learnings from my journey (more to explore and excited for what lies ahead) in a new Substack article – a growing collection of wisdom! ✨ Check it out here: https://lnkd.in/ewjsB5xc I believe everyone has valuable insights to share. 💡 Feel free to post in the comments any quotes or learnings that significantly impacted your life or career Let's inspire and empower each other! 👇 #quotes #wisdom P.S: This Substack article will be updated regularly with new insights, so be sure to subscribe to stay in the loop! 🚀

11

1 Comment

🚀In my newest Medium article, I’ll be diving into the world of text input dialogs with Avalonia UI. Join me as we explore how you can create modern and effective text input dialogs that enhance the user experience. GitHub: https://lnkd.in/dtrjPRhz Medium: https://lnkd.in/d4T_PFRT Thanks! #AvaloniaUI #CrossPlatform

26

2 Comments