Xiao Li

🌟 From Raw Data to Behavioral Predictions—Effortlessly What if AI deployment didn’t require months of preparation? With BaseModel.ai now on the Snowflake Marketplace, you can transform behavioral data into predictions in just a few clicks. ✅ Rapid Deployment: Build and deploy models in a week.  ✅ Scalable Flexibility: Create multiple models for diverse use cases.  ✅ Privacy Guaranteed: Data never leaves the secure Snowflake environment. Say goodbye to complexity and hello to actionable insights. AI for behavioral data has never been this simple. 👉 Try BaseModel.ai Today  #AI #MachineLearning #BehavioralData #BaseModel #Snowflake #Synerise 

39

4 Comments

"Build in primitives" -Andy Jassy, #Amazon CEO https://lnkd.in/gHMYt7NX Happen to take the time to read this year's Annual investor letter from Andy Jassy. Deep in 2nd half, it talks about how Amazon started thinking about "building in primitives" and how it has been game-changing for them and their customers. Probably old news for some, but eye-opening for me. Definitely a good read!

2

"Over 25% of all new code at Google is now generated by AI", according to CEO Sundar Pichai in today's earnings call. But is this shift aimed at speeding up the development of complex software, or simply cutting engineering costs? One potential indicator will be if Google reduces headcount by 25% while boosting revenue per employee. Here are some details behind this productivity metric in Google's own research page . Majority of this '25% code' is in the area of code completion, not in code creation. "LLM-based inline code completion is the most popular application of AI applied to software development..." Watch the video to understand further on this metric.

16

3 Comments

Google announced a new Enterprise version of Notebook LLM among their rush of big announcements last week. Looks like an interesting Enterprise offering which supports up to 300 sources per notebook - a reasonable cutoff for department level knowledge retrieval use cases. This is a big improvement over the OpenAI Enterprise GPT solution which caps at 20. Still not sure about the podcast feature. Feels a little gimmicky? Definitely recommend checking out Notebook LLM consumer offering if you haven't yet. Pretty nice tool. #Google #OpenAI #GenAI

16

2 Comments

LLMs are going to be commoditized akin to cloud services in the early 2010s. Big companies and startups alike are going to start building agents on top of these offerings. Whoever can solve the latency, guardrails and security issues will come out on top.

11

Will be really interesting to check how mixed Agentforce agents for Sales, Service, Marketing, Commerce, etc. work together that would balance the utilization of cloud computing resources, or instead influence cloud providers’ computing, memory, networking resource configurations.

15

1 Comment

Recently, I introduced Ava, a custom GPT to help Engineering Managers with planning, strategy, hiring, etc. I ran an experiment with Ava to role-play an interview. 💬 Ava took on the role of an interviewer and an EM candidate. It generated questions, answered them, and graded the responses. It was like running a real interview, but with a AI doing most of the heavy lifting! While the answers weren’t as detailed as you'd get in a real interview, the results were interesting 👉 https://lnkd.in/ec488vy6

5

2 Comments

Shameless rip from Austin Parker’s BlueSky. But this whole thread is just too good! If you are passionate about #observability you’ll love this! Find the rest here! https://lnkd.in/gsvS_Jpz “”” 1. Metrics are good, actually, but most of you use them very poorly. … 4. there are no silver bullets when it comes to telemetry. you're trading between speed of access, data resolution, and retention … 5. you are not going to build a better datadog on-prem, they have a twelve year and twenty billion dollar head start … 18. the actual goal of observability is to have the data you need to find the answer rather than to have the data that contains the answer … 32. if you have thousands of dashboards, you have no useful ones … 44. single panes of glass are for glass cleaner, not o11y “””

40

5 Comments

Hey everyone, I am an investor and advisor to Katanemo.... They are reinventing the infrastructure primitives for GenAI and launched Arch - https://lnkd.in/gCrbbkEV - an intelligent prompt gateway. Give the project a look on Github, and star the project if it resonates

88

3 Comments

>14K KIDS + >10K WOMEN = >65% of the dead = U.S. funded GENOCIDE 👉🏿 Please listen to U. of Chicago Prof. John Mearsheimer explain how Israel is dragging us into a war with Iran, which could easily spiral into WW3. He says Israel believes "the bigger the war, the greater the opportunity for ethnic cleansing” & has always used war as a cover for ethnic cleansing: lnkd.in/g9tDWs2i 👉🏿 What would you do if someone stole or demolished your house & livelihood, killed some of your family, & left the rest of you to live in the squalor of a refugee camp? That is what Israel did to over 750K Palestinians in 1948, over 300K in 1967, over 250K since 1967, & MORE EVERY YEAR. 👉🏿 Most of the families in Gaza owned homes & farms where Israeli cities & towns now stand, which Israel literally stole from them. Many even still have their front door keys. AND NOW ISRAEL IS DOING IT TO THEM FOR THE 2ND TIME. 👉🏿 Jewish-Israeli sources: 1) Prof. Ilan Pappé, Historian: lnkd.in/gFBK6sPB 2) Miko Peled, ex-IDF soldier, son of an Israeli general & grandson of one of Israel’s founders: lnkd.in/geiy3_5s 👉🏿 The root cause of this horrible conflict is that Israeli Zionists actually believe they own every Palestinian home & farm, & have the right to kill indigenous Palestinian families who won’t leave, FOR THE CRIME OF NOT BEING JEWISH. 👉🏿 Here is renowned Jewish-American Prof. Noam Chomsky explaining how ridiculous the Zionist ownership claim is: lnkd.in/gtHPPHUz Palestinians have been there for thousands of years, never left, & their genetics proves it. At some point, their Canaanite & Jewish ancestors simply converted to Christianity &/or Islam. So why would that mean Jewish immigrants from Europe & the U.S., who have never even been there, own all of their homes & property? 👉🏿 And this article explains how Israel has always been able to beat Hamas & ensure its security WITHOUT KILLING CHILDREN. Israel ignores this option because its goal is not to beat Hamas, but rather to keep using them & similar groups as a pretext to continue ethnically cleansing more of historic Palestine. That’s why Israel helped strengthen Hamas for years & continues using indiscriminate brutality against all Palestinians even though it only helps Hamas & similar groups recruit. Most of the 30K Hamas fighters today were orphaned by Israeli bombing over the past 20 years, & Israel just created around 30K more orphans since 10/7: lnkd.in/gY98T9mc 👉🏿 I also cry for the 38 Israeli kids killed on 10/7, but this horrible conflict did NOT start on that day & there was no ceasefire. The proof is the nearly 1400 Palestinian kids killed by Israeli soldiers & settlers between 2008 & Oct 2023, including 43 kids during the 9 months just before 10/7 (all details on BTselem.org) 👉🏿 And Israel has also imposed a brutal 57 year occupation/apartheid on Palestinians, & a brutal blockade of Gaza since 2005. Both illegal, immoral, & ongoing acts of war. 👉🏿 All paid for with OUR TAX DOLLARS

5

1 Comment

So many things to unpack in the wake of another company squeezing their open source users towards licensing. Thinking of the current OSS climate and the lack of participation of companies generally as well as not being able to endorse using a solution fearing that a company will pull the rug out from under you once you're fully integrated with a solution. https://lnkd.in/gVpME4Uz

2

1 Comment

Check out insights from @Incorta and @GoogleCloudPartners about how 🤖 AI solutions & ML models are only as good as the data that fuels them. If your operational data is old, inaccurate, or missing, your next-gen tech could rely on false assumptions. Get the full ebook to discover how Incorta works together with solutions like Google BigQuery and Google Cortex Cloud Framework to help you: ✅ Power ML models and AI solutions with live, detailed data ✅ Effectively ingest, correlate, and apply data at scale to support analysis ✅ Get a holistic view of data from across the enterprise for more accurate analysis & better decision-making #AI #ML #GenerativeAI #data #analytics #OperationalLakehouse #Incorta #GoogleBigQuery #cloud #GoogleCortexCloudFramework #GoogleCloudPartners

3

I read a lot of hyped up click bait AI garbage posts on LinkedIn from people with 4 years GenAI experience and a previous background in Bitcoin and Web 3.0. Having spent a little time at the coal face trying to implement applied innovation of AI at scale for the worlds largest and most successful bank as a Technical Product Manager this is the best article I have read on the subject. #GenAi #AppliedAI #AItransformation

9

2 Comments

This is very interesting! I usually try to follow this lesson. The one item I can note here is that you need a team that will go back and fix the issue properly!

1

1 Comment

Marking a decade of Amazon Cognito, with a big bang! Beyond thrilled to announce the launch of multiple new capabilities to help developers setup seamless and secure authentication within minutes using Cognito! #aws #cognito #preInvent We announced: (1) a #nocode visual editor to personalize and localize sign-up and login experience for applications, (2) low-friction and secure sign-in experience with passwordless login, (3) a revamped #developer experience, and (4) new product packages with the pricing of our Plus package, allowing you to save up to 60% of monthly spend vs previously using our advanced security features! https://lnkd.in/gyTseR8n

82

9 Comments

𝗖𝘂𝗿𝗶𝗼𝘂𝘀 𝗮𝗯𝗼𝘂𝘁 𝘂𝘀𝗶𝗻𝗴 𝘁𝗼𝗽 𝗔𝗜 𝗺𝗼𝗱𝗲𝗹𝘀 𝗳𝗿𝗼𝗺 𝗠𝗲𝘁𝗮, 𝗠𝗶𝗰𝗿𝗼𝘀𝗼𝗳𝘁, 𝗮𝗻𝗱 𝗚𝗼𝗼𝗴𝗹𝗲—𝗳𝗼𝗿 𝗳𝗿𝗲𝗲? 🤖🧠 If you've been wanting to dive into the latest AI models but don't have time to set them all up, I’ve got some good news. 𝗛𝗲𝗿𝗲 𝗮𝗿𝗲 𝘀𝗼𝗺𝗲 𝗽𝗼𝘄𝗲𝗿𝗳𝘂𝗹 𝗺𝗼𝗱𝗲𝗹𝘀 𝘆𝗼𝘂 𝗰𝗮𝗻 𝘁𝗿𝘆: 𝗟𝗹𝗮𝗺𝗮 𝟯.𝟭 from Meta – Known for its ability to generate human-like text, Llama 3.1 is fast and efficient, so you can experiment with advanced AI without needing a super-powerful computer. https://llama.meta.com/ 𝗣𝗵𝗶 𝟯 from Microsoft – This model is all about speed and scalability. It’s great for anyone looking to enhance automation and quickly process data without getting bogged down. https://lnkd.in/gPwQt-jJ 𝗠𝗶𝘀𝘁𝗿𝗮𝗹 – This up-and-coming model is gaining attention for being agile and quick. It’s great for playing around with text and creative projects—perfect if you like to test and iterate fast. https://mistral.ai/ 𝗚𝗲𝗺𝗺𝗮 𝟮 from Google – Google’s Gemma 2 is super sharp when it comes to understanding tricky queries and generating detailed responses. It’s a solid choice if you’re building tools that need smart, context-aware responses. https://lnkd.in/gkte-WtF 𝗕𝘂𝘁 𝗵𝗲𝗿𝗲’𝘀 𝘁𝗵𝗲 𝗰𝗼𝗼𝗹 𝗽𝗮𝗿𝘁: 𝗬𝗼𝘂 𝗱𝗼𝗻’𝘁 𝗵𝗮𝘃𝗲 𝘁𝗼 𝘀𝗲𝘁 𝘁𝗵𝗲𝗺 𝘂𝗽 𝗼𝗻𝗲 𝗯𝘆 𝗼𝗻𝗲! There’s a package called Ollama that includes all these models, so you can try them out easily. And if you want to integrate them into your own projects, you can use the OllamaJS library to get started. https://ollama.com/ 𝗪𝗵𝘆 𝗯𝗼𝘁𝗵𝗲𝗿 𝘄𝗶𝘁𝗵 𝘁𝗵𝗲𝘀𝗲 𝗳𝗿𝗲𝗲 𝗺𝗼𝗱𝗲𝗹𝘀 𝗶𝗻𝘀𝘁𝗲𝗮𝗱 𝗼𝗳 𝗽𝗮𝘆𝗶𝗻𝗴 𝗳𝗼𝗿 𝘀𝗼𝗺𝗲𝘁𝗵𝗶𝗻𝗴 𝗲𝗹𝘀𝗲? 𝟭. 𝗟𝗲𝗮𝗿𝗻 𝗯𝘆 𝗱𝗼𝗶𝗻𝗴: There’s no better way to pick up skills than by experimenting yourself. 𝟮. 𝗦𝗮𝘃𝗲 𝗺𝗼𝗻𝗲𝘆: As long as your hardware can handle it, you don’t need to spend big bucks on premium AI services. 𝟯. 𝗞𝗲𝗲𝗽 𝗰𝗼𝗻𝘁𝗿𝗼𝗹 𝗼𝗳 𝘆𝗼𝘂𝗿 𝗱𝗮𝘁𝗮: You don’t have to send your data to some unknown service—you can experiment locally and privately. Once you get the hang of it, you’ll see just how much fun it is to explore these models. And who knows? You might discover new ways to use them that could lead to something bigger. #AI #LLM

8

1 Comment