Alex Xu

🤖 Creating a fully autonomous AI developer to write and maintain code is still the holy grail of software development. There are many brilliant teams currently working on this, and each of them have their own unique approach. The team at Cosine have made a massive step towards making this a reality. “Cosine achieves state-of-the-art results on the SWE-bench benchmark” They have achieved a score of 43.8% on SWE Bench, which was verified by OpenAI. What's even more incredible is they have managed to do this with a lean team and a relatively limited budget. Though it’s important to note Cosine isn’t looking to replace human developers entirely, but rather augmented them with human-like assistants that they can collaborate with on any kind of coding task. Yang Li + Sam Stenner + Alistair Pullen Read more about their collaboration with OpenAI 👇 https://lnkd.in/gu7xXVPn

21

Everyone's talking about “founder mode,” but as it gains traction, founders will start using it as an excuse for poor management. Founder mode isn’t a license to run a company at 5,000 the same way you ran it at 5. Jensen Huang, CEO of NVIDIA, exemplifies what I like to call “scaled founder mode”. Scaled founder mode is being highly involved in the most important things, that only you can do to drive the company forward. Nvidia is one of the most successful tech companies in the world, with a $3 trillion market cap. And Huang, with a 31+ year tenure, is the longest-serving CEO in the industry. He’s a highly effective manager. And he’s still very much in founder mode. When interviewed by Stripe's Patrick Collison, Huang said “CEOs are pinch hitters. We should be working on the things that nobody else can or nobody else is.” Nvidia’s leadership team of 60 reports directly to Huang. But he doesn’t do 1:1s. This way, the team of 60 is relatively flat, with no silos and no privileged access to information. Not saying that killing 1:1s is the answer to speedy execution. But Nvidia found that this way, everyone contributes to solving problems with equal access to important info. And because Huang sees his primary role as being a custodian of the culture, he eats in the cafeteria with employees every single day. At almost 30k employees, you can’t get much more “in the details” than that.

26

3 Comments

Ilya Sutskever has just launched a new AI venture called Safe Superintelligence (SSI), aimed at developing safe and powerful superintelligence. This comes one month after he left OpenAI. SSI is co-founded by Sutskever, former Y Combinator partner Daniel Gross, and ex-OpenAI engineer Daniel Levy. The sole focus of SSI is to create beneficial superintelligent AI, which Sutskever believes could emerge within a decade. SSI will be a pure research organization, insulated from short-term business pressures. #ai

2

Some responses to this article have taken the point of view that muffling the angry customer calls is itself harmful. What do you think? Here is my take: Drawing on my background in business operations, where navigating crises is part of the job, I've learned that anger and extreme emotions rarely lead to effective solutions. The truth is, understanding the severity of a problem doesn't require being yelled at. Discovering and resolving issues can be achieved without resorting to verbal abuse or harassment. In fact, such negative behaviors only hinder the problem-solving process. What responsible employer would ever condone a customer mistreating their employee? I've found myself in the role of a frustrated customer on several occasions. Yet, I understand that the service agent is seldom the cause of the issue. Therefore, I refrain from directing any anger towards them. This doesn't make me a saint—it's about basic respect and civility toward another person.

1

Ophelia Zhang Dalsgaard and my boss Henrik Høegh on stage at #KCDDenmark talking about #DevX and #PlatformEngineering is a Venn diagram. 👏 There are overlaps. But actually it’s good old user experience (UX), they coined it PUX (Platform UX). The point is that there’s much more to DevX than what a platform can solve. Developers are not necessarily more productive if they are stuck in endless meetings, have redundant and meaningless processes, etc. Achieving flow for developers is for the whole organization to solve, because it’s just good and obvious business value!

38

1 Comment

My bets where AI won't replace humans: 1) Human connection: Like the one that many of us just had at Config. 2) Shared human experience: When you look at another person's eye and you know you've been through a similar joyful and painful journey. 3) Real stories behind art and culture: I listen to music, and I like paintings not just because they feel good, but because there's a real human story behind it that speaks back to 2. 4) Ingenious net new creative fields: AI can use past things to create something new by mixing them, but it's mostly always can't really create something completely new and unseen. 5) What's your bet?

34

4 Comments

The revoultion in Storytelling coming to us via Generative AI is two-fold. One, we are soon to be able to assemble small, and very talented creative teams (writers, artists, directors, and actors) and create vast universes and intricate stories that while still driven from their center by human hearts, minds, and hands, have the ability to explode into being and even evolve over time at the highest fidelity and in myriad or even blended styles. Two, these narratives will become increasingly personalized, immersive, but still social shared experiences that sustain engagement in ways that give each of us what we love most without compromising the integrity of the storytelling universe, succumbing to the weight of insurmountable costs, or being hindered by the production timeline realities of today. It will be a disruptive shift, but one that can be ultimately democratizing in ways that will eventually encourage and empower artists again and give rise to new talents and new stories of all kinds because the gap between the audience and the creator will close to the point of intimacy. #media #generativeai #storytelling #ai #artists #streaming

11

Anthropic's latest Claude AI model is taking human-AI interaction to a whole new level! 🖥️🤖 The new Claude 3.5 Sonnet can now use computers just like a human would - by looking at the screen, moving the cursor, clicking buttons, and typing text. This is a major leap forward in AI capabilities. Key features: Operates based on screenshots, emulating human computer use Can perform complex multi-step tasks across different applications Self-corrects and retries when encountering obstacles2 Accessible via API for developers to build with Why it's game-changing: This opens up incredible possibilities for AI assistance in data entry, form filling, software testing, and automating repetitive computer tasks13. Imagine having an AI assistant that can seamlessly work alongside you, handling tedious computer work while you focus on higher-level tasks. While still in beta, this technology demonstrates how AI is rapidly advancing toward more human-like interaction with digital interfaces. As it develops further, it could revolutionize how we work with computers and boost productivity across industries. The future of human-AI collaboration just got a lot more exciting! What do you think about this new capability? #AIInnovation #FutureOfWork Here is a video about it: https://lnkd.in/gh8aNYEu

6

Curious and slightly unsettling experience with chatGPT during a novel review processl (it has been a long time since the publication of my last fiction book!). I uploaded 4 chapters using the premium version and requested summaries of the main characters' names, surnames, and descriptions. One character intentionally had no surname, but chatGPT provided one. After pointing out the error, chatGPT apologized, acknowledged the mistake, and offered a different surname. When I inquired about the repeated incorrect surnames, chatGPT explained its reasoning: Certainly! The reason for the earlier errors with the characters' surnames is a mix of factors: Initial confusion in searching the document: When I tried to retrieve the names from the original document, the search didn’t extract the correct information, leading me to create plausible but incorrect surnames instead of relying on the actual text. Autogenerated names: In cases where I couldn’t locate the correct surname, I defaulted to generating surnames that seemed fitting for the context or the character’s first name. This was a mistake, as I should have used the accurate details from the text. Imprecise searches: My early attempts at searching the document didn’t yield the right results, so I made assumptions instead of double-checking the content properly. #chatGPT #AI

23

2 Comments

On April 18th, just last week, Meta unveiled Llama 3, lauded as "the most capable openly available LLM to date". This remarkable achievement by the Meta team comes hot on the heels of Llama 2's release last July. Notably, the new models surpass their predecessors and competing offerings from other providers in terms of performance. Meta's commitment to open sourcing these powerful models while prioritizing model safety and responsible usage is commendable. (https://lnkd.in/ghv3RJRz) Yesterday, Microsoft introduced Phi-3 Mini, a member of the Phi-3 family of models, promoted as "the most capable and cost-effective small language models available". According to Microsoft, the key to achieve high performance in such a small package lies in the quality of the training data. The model has been open sourced and is now available on Ollama and Hugging Face. (https://lnkd.in/gKgSmrBn)   These compact yet powerful models serve as invaluable tools for creators eager to bring their visions to life or simply experiment with cutting-edge technology, particularly for those with limited access to resources.   Shortly after the release of Llama 3, reports began surfacing of enthusiasts running the 8B model on a Raspberry Pi 5 equipped with just 8GB of RAM—a modest $80 single-board computer. (https://lnkd.in/gbeQPhQ2) While the performance understandably reflects the limitations of such a platform, the fact that the model runs and generates an output is nothing short of magical. Similarly, the Phi-3 Mini model can run on ubiquitous devices such as smartphones.   Reflecting on my own journey, I recall the pivotal role that early access to a low-cost computer played in shaping my passion for coding. The gift of an 8-bit Commodore 64 home computer ignited my curiosity at a young age. Later, with the guidance of an outstanding Computer Science teaching staff at my high school IISS Marconi-Hack Bari, I embarked on ambitious projects like a networked multiplayer version of Battleship written in Turbo Pascal and 8086 Assembly, complete with custom sprites and lo-res graphics. These formative experiences paved the way for my career, eventually leading me to various engineering and leadership roles.   It all began with a simple, low-cost, modestly capable computer used for experimentation—a gateway to discovering my life's passion.   I encourage educators and parents alike to consider setting up similar low-cost experimentation environments. By introducing younger generations to open-source AI technologies, we can nurture their creativity and help them uncover their true passions. It's not just a weekend project—it's an investment in their future and the future of innovation and discovery. #education #youth #AI #LLM #future #innovation

24

[Sorry, there was a mistake on the project name in our new large multimodal model post. We did reformat the multimodal project as XGen-MM as part of our XGen initiative. We are now reposting the project with the updated name. Thanks.] We introduce XGen-MM, a series of large multimodal models (LMMs) developed by Salesforce AI Research. XGen-MM is a new SOTA model under 5B on few-shot learning and multimodal benchmarks. Check our first HF release at https://lnkd.in/gHYrmSmm, and stay tuned for the coming technical report and future releases! Key Takeaways: XGen-MM is trained at scale on high-quality image caption data and image-text interleaved data. XGen-MM exhibits strong few-shot learning capability as pre-trained base model and impressive visual chat ability after SFT. XGen-MM supports more efficient token usage compared to methods like LLaVA (more than 5x less!).

81

Looking forward to sharing how we at basys.ai are leveraging Generative AI to tackle healthcare’s biggest challenges. PS: Also looking forward to sporting my basys vest in the #BayArea PPS: What I’m not looking forward to is the second cross-country flight in less than a week. #GenAI #BasysAI #GenerativeAI

20

how do you get jensen NVIDIA to buy your ai start up. you make a Run:ai for it ;) PayPal. we talk about peter theil, elon musk, the rest of the gang as being exceptional. frankly i don't believe that except for one massive differentiator. perhaps it was theil who was aware that solving payment transactions for sellers, buyers on the wildly success eBay - the auction house at your finger tips. the data was there. 1000s of sellers. millies of products. cash flow. interrupted. so solve the interruption. and do so to the point where eBay must buy you. what's so sophisticated about this is that eBay already had all the data on PayPal to the penny. it was a matter of hanging on. and they did. a bit of luck. a bit of timing. an extraordinary exit. and suddenly the gang had some money in their back pocket jeans. this 700 or 800 milly deal between NVIDIA and Run:ai smells the same. Omri Geller came up with a math solution. to matrix multiplication. NVIDIA gpu resources reconfigured. via software. to optimize the optimization of ai models. in terms of timing this deal also has OpenAI's sama to thank. as he jetset 'round the world speaking on the great worry of safety with ai to governments the politicians began to freak out. coincidently many elections were up for grabs. what? with this ai my competition may take over. oooops. better get a handle on this. this fomo was also evident with NVIDIA's stock rising faster than any other in recorded history of market capitalism. trillys in market cap over 2 years. unique. there were others in the mag7 who also benefited from a rising investment tide. Run:ai just kept on running. and their orchestration became that much more valuable. in fact doubled just because of the stock prices. demand. elon creating xAI. setting himself up for a 1 milly NVIDIA gpu cluster. the industry is aware of existing limitations. Run:ai rather than whining, they're win. ning. sadly i was reminded that places like Vector Institute or Mila - Quebec Artificial Intelligence Institute or Alberta Machine Intelligence Institute (Amii) or other global research centres like ETH Zürich University of Oxford prob had the same level of talent yet definitely didn't have the vision, the investment capital, or attitude to do something about it. this is what we mean by ai being a winner takes all game. this orchestration being solved and now to be open sourced means there is no 2nd place. 3 millys seed TLV Partners 10 millys a SCAPITALVC TLV Partners Sequoia Capital 30 millys TLV Partners Insight Partners 75 millys TLV Partners SCAPITALVC Sequoia Capital ---- 118 millys 700 - 800 millys from jensen NVIDIA. bought to optimize all NVIDIA gpus from the get go ++. source: https://lnkd.in/eHJQ8nWK source: https://lnkd.in/etuCyEfJ source; < . >

15

4 Comments

MOST Gen AI POCs will fail to make it to production. Why? 1. We have had some potential clients expecting us to build POCs for free ("You can put our logo on your slides, but we won't pay") - frankly if you're not willing to pay for a product, you haven't even given it enough thought or consideration. It's just wasting everyone's time. And Gen AI is not a cheap solution, especially when it's custom build. 2. There is no clear understanding of 'cost of compute'. "Let's just do a POC now, and we will figure deployment later". There's a reason why NVIDIA is a 3 Trillion $ co - there's a chip demand-supply mismatch and cost of a product (compute) in a a supply restricted market is going to stay elevated in short run. No clear deployment plan post POC is an another brilliant dead end. 3. No Data strategy - you just can't use ALL of your data on one single LLM. One of the biggest problems we have seen (and are addressing for our clients) is using PII data in Gen AI workflows. People we meet in Insurance, Banking, family office etc tell us there is more data they can't put onto Gen AI workflows, than the one they use. Perfunctory data strategy, or using simulated data to push through a POC won't get you anywhere meaningful. 4. One LLM to rule over all mistake - Way too often we come across "Lets just use the most powerful LLM and do all of this". When the fact is, some LLMs are better for some use cases you are trying to solve. Keep aside the compute cost, deployment flexibility, ownership of data etc. LLMs are tools - just like hammers and screw drivers. You can, if you wish to, hammer a nail with a screw driver, but… 5. Too many POCs - how many POCs are too many? We met someone who showed us a deck of 58 POC ideas that their company is working through (I am understating that number). I was left wondering, isn't it better to just get team leaders from different functions into a room, have a simple discussion with a whiteboard and list what problems would they like to eliminate? 58+ POCs is just lazy - the list was built by prompting chatgpt - "give me 5 dozen Gen AI POC ideas for an Bank" Gen AI is not magic. It won't make all your problems disappear over a weekend, it won't suddenly make your sales guys super heroes, or get your costs down by 50% by the time you go on your Christmas holidays. It is a super powerful tool for sure, and instead of wasting time and money on 5 dozen POCs, it might just be better to 3-5 solid ones with a clarity of how to execute till the end.

19

2 Comments

🚀 New Updates to Gemini Models! 📢 Two new production-ready Gemini models are here: Gemini-1.5-Pro-002 and Gemini-1.5-Flash-002. Key highlights: 💰 Over 50% price reduction on 1.5 Pro for prompts <128K 📈 2-3x higher rate limits for 1.5 Flash & Pro ⚡ 2x faster outputs and 3x lower latency 📊 Significant improvements in math, long context, & vision tasks 🧠 Enhanced performance in code generation and visual understanding They're already available on Google AI Studio & Vertex AI. More on this: https://lnkd.in/dwrCM957 #Gemini #AI

18

NVIDIA has a clear “unfair advantage” in the market right now. They are ahead of the game in so many fronts, they sit on the hardware every company working with AI needs and is the preferred partner/CVC for any AI startup at the moment. If I was Jensen Huang, I would issue 10% new stock and raise a $300bn war chest at the same time as they are doing the 1/10 split. I would take $10bn of that money, split it into 40k tickets of $250k and fund basically any new novel AI project (no ChatGPT wrappers). This way NVIDIA would have a stake in ~75% of all AI related startups and the massive amount of talent those companies attract. They would ultimately become unstoppable while still having $290bn in the war chest 🤯

23

1 Comment

proximity bias in AI, fueled by "ecosystems dominance" SV is great, but super-concentration of SV as well as near monopoly of ecosystems like iOS, Android and Facebook/WA - the world's best experiences will always come to rich US cities first to the people who are building it. Because the developers are all there and it's easier to build great products for themselves. Apple folks, by large are living in a 30 miles radius around Cupertino - so US101 traffic will be the first to be predicted, english will prevail and typical bay-area centric apps will be the first to have actual good intelligence. Surprisingly, even a super exclusive Greater China Market didn't stop Apple from building a moat there. Although the power of Apple there is much small (mainly hardware advantages rather than software - Apple pay is weak, no Apple TV/Music strength, macbook adoption is much smaller...) Regulation might be the only hope here for regions to jump ahead and break the moat, but it'll be real messy. Until then, let's first marvel at how great soccer dad/mom's saturdays would be with the help of Apple AI. https://lnkd.in/gDkQnfZc

6