Antoine Dusséaux

Why SanFran as city is unique + Europe can’t compete ? 300 000 millionaires & 80 Billionaires Even without capital raising, the probability of good knowledge transfer about how to build substantial (software) businesses seems to be working better on an order magnitude in that city, than in Europe as a whole

2

The field of LLMs is getting more and more interesting. So many real world use cases are emerging, making a different shift in digital transformation. Stay tuned for more details. #LLMs #digitaltransformation

9

🧩 One thing we've noticed with #VentureIQ technology is how well it can distill and communicate complex technical material 🧩 Whether it's a #deeptech venture, a novel #AI application, or even a patent submission, we can quickly distill the business and strategic implications for investors...and sometimes even the founders/PIs themselves! 🤯 Based on our work so far, the tech can handle a wide array of research areas and technological sectors -- anything from blockchain algorithms to vaccine development. It handles the nuance of the field, the specific terminology, and extracts the key concepts which matter for venture-building. 🤓 To demonstrate, I took a patent I helped write with Eric Laber on reinforcement learning. I ran it through #Ventureiq to build a commercial assessment (key differentiators, target markets). Then for fun I had #NotebookLM make a podcast out of it. I've uploaded the result it here (DM me if you'd like the actual report): https://lnkd.in/eiTejsny This is more than just automation or streamlining operations - it's about unlocking clarity and insight when it was previously unattainable, and the ability to do it over and over again no matter the sector or vertical. Interested in more info? DM me so we can chat.

11

2 Comments

Vishal Sachdev highlights the strategic integration of open source and proprietary tech in architecting tech stacks, developer ecosystems and resulting business models. The world class companies effectively balance value commoditization in open source and value capture in proprietary tech. This is the strategic challenge for companies operating in #blockchain and #AI. I believe those who operate at the intersection of blockchain *and* AI stand to win this strategic battle...

3

1 Comment

Some great observations and clarifications here by Francesco Perticarari dispelling myths from facts about #DeepTech #Investments using data.

5

🤖 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

Lean agents frameworks are the new hot thing - recently Anthropic, and now Hugging Face. On the surface, less is more, however, there are nuances here: 1) Early agent frameworks focused heavily on orchestration, tool use, structured output, and code execution - capabilities now increasingly built into foundation models themselves. On the one hand, tight vertical integration is great, on the other, so is competition... 2) Space is still very early with many low-ish hanging fruits. I expect the frameworks to focus on these: - More sophisticated memory architectures (including x-session), integrating episodic (specific events), semantic (general knowledge), and procedural (how to do things) memory, leveraging smart abstractions and summarization. - More advanced planning and reasoning - like hierarchical task decomposition, Look-Ahead & Simulation, and integrated Program Synthesis and Discrete Reasoning modules - Better Error detection, self correction strategies, and learning from past mistakes and successes - Better tool understanding and selection - More sophisticated inter-agent communication protocols, dynamic role negotiation systems, and conflict resolution mechanisms.

9

🚀🔍 A great analysis by Index Ventures on Vertical AI companies as part of the Application Layer of the AI stack. "With the AI Platform Shift upon us, we believe that the next logical iteration of vertical SaaS will be Vertical AI – vertically-focused AI platforms, bundled alongside workflow SaaS, built on top of models which have been uniquely trained on industry-specific datasets." The few things that will differentiate the winners in the race, as mentioned in the article:  - access to proprietary industry data,  - effectively train large language models against those datasets,  - package those models via applications,  - deliver tremendous utility with fast time-to-value for customers, At Neurolabs, we care a lot about the time-to-value, which is why we’ve built these exact steps for our Image recognition solution.✅ With speed of execution & user experience at core, we go a step further and train our large vision models with synthetic data, delivering an even faster onboarding experience for our CPG customers & partners. 🤝 If you're facing challenges with your IR projects, let’s talk. There's no need to rely on outdated methods. Synthetic IR is here to propel us into the future! 🚀 #syntheticimagerecognition #syntheticdata #computervision #imagerecognition #cpg #retailexecution  

16

1 Comment

The most important fundraising KPI: 1st pitch → 2nd meeting conversion rate Is yours low? (< 30-50%) Something is wrong. Here's how to diagnose your issue and fix it: There are two types of potential problems... 1️⃣ Investor targeting 2️⃣ Pitching 𝐈𝐍𝐕𝐄𝐒𝐓𝐎𝐑 𝐓𝐀𝐑𝐆𝐄𝐓𝐈𝐍𝐆: The most common issue is targeting the wrong investors. Usually you're reaching out to investors that focus on a different stage or industry. Do some research on the investors that didn't convert to a second meeting. The most important resource: their portfolio page. Ignore the "about" section, or their blog. That's mostly about marketing to potential investors in their fund. Instead, look at what they've actually invested in and the stage at which they invested (Pre-Seed, Seed, Series A, etc.) Remember: trust the portfolio, not the thesis. __ The other main issue is unclear outbound messaging. Investors think you do one thing (industry, business model, tech, etc.) but you actually do something else. I recommend sending your outbound email template to 5+ friends that don't know anything about your startup. Ask them: • What kinds of customers do you think we serve? • What benefits do you think we provide for them? • What product features do you think we have? • What technologies do you think we use? If more 20% of your friends answer these questions incorrectly, you need to write new outreach emails. 𝐏𝐈𝐓𝐂𝐇𝐈𝐍𝐆: The most common issue is just that your pitch is weak. Usually because you aren't telling a story. You're just sharing disconnected data. The best pitches are stories that foreshadow the only possible outcome: your startup's tremendous success. It feels inevitable. The easiest test to figure out if you're telling a story: try connecting the main idea of each of your slides with "but," "so" or one of their synonyms. If the natural link between slides is "and" or if you can move the order around, you don't have a story. Chronology matters in storytelling. Go back to the drawing board and try to write your inevitable success story. __ Maybe your overall pitch isn't weak. But VCs may still be spotting a red flag when you pitch. It turns them off right away. If you think this might be your issue, it's important to start asking questions. Most pitches have time for you to ask questions at the end. Here's my go-to to diagnose this issue: "What would you need to see in order to get excited about investing in <<STARTUP>>?" If they tell you a primary KPI that you need to improve, then you've probably identified the issue. The other possibility: They try to avoid giving you a specific answer. If they're evading your question, you NEED to dig in to get to the bottom of it. Your fundraising success may depend on it. —— Was this helpful? 👍 like and ♻️ repost it to help other founders!

55

27 Comments

Have you heard about Tools for Humanity and Worldcoin? Another grand idea by Sam Altman - even more dangerous than GenAI as it does not rely on AI at all but on brut force control. Basically, he scares you that one day no one will be able to tell the difference between you and a robot, unless……..! Unless, of coures, if you let him scan your eye and this scan becomes the Alrman issued world passport. But there is more to it. He pays you for the scan with World coins. You even make some money from the scheme. The more eyes are scanned, the more the value of the workdcoins grows. Brilliant. But is it by any chance that he started the eye scanning business in Nigeria and Kenya? No. He gets cheep eye scans that in turn raise the value worldcoins. He can now pay the rest of the expenses without fund raising. It is not by chance that world governments are up in arms against this. There are so many issues and the plan is so grand - creating and owning the world identity company. Did he come up with this idea just like that? And did everyone else thought that he was crazy except the gullible 26 year old who he hired to run the company. I doubt it. There is a definite political agenda here that rivals in ambition those of Napoleon, Hitler, Stalin and Putin together. Great ambissions like this one lead to great misery and suffering if not stoped on time. https://lnkd.in/eejCqABc

7

1 Comment

Everyone is focusing on how powerful models are. But everyone might be thinking wrong. Why?  In a post last week, I was discussing the decision of Meta to not rollout Llama 3 in EU. The decision was mostly driven by a conflict on the origin of the data used to train the model. And this is the most important part : the data used to train the model. For must of us, the problem when using an LLM isn’t how powerful the model is but the quality of data To compare it, if you give expired food to the best chef in town, you’ll still end up sick. On the other hand, it has been showed, especially in financial forecasting use cases that a simple model can have amazing results with the good data. So next time you’re looking at an AI product or startup to invest-in try dig as fast you can on how and where they get their data from.  Start-ups that have an intelligent way to have or a create a proprietary dataset will have a huge advantage on those relying on API and third parties. Now more than ever data is the oil of the 21st Century. 

45

1 Comment

Its tempting to say AI is the new oil. But oil is what is burnt. The engine burns the oil. Data and Information continue to be the Oil. We are in that same Information Age as theorised by Manuel Castells et. al. What's new is that we have just invented our Engine! And that's GenAI. Now we can put the Data to use... #genaiadoption #nlightn

6

Attached is NVDA's price chart for the last year - up c.175% (up 2500%+ in L5Y period!!) 🚀 I nonetheless have conviction that it currently remains undervalued... Among other things, due to strong competitive positioning as a GPU-maker, AI-driven demand fuelled by investor and consumer interest, and moats it has built. This is mirrored by equity analysts – GS’ latest price target was $800, BofA’s $1,500. I do believe the growth is capped, and driven in equal measure by under-supply and bubble-like demand as it is by fundamentals. Even so, Nvidia remains an attractive medium term investment. ⚖️ Demand – Supply Dynamics: AI is a generational innovation, and agents /apps are already upending SaaS, or at least so think VCs who deployed $64.6B in AI start-ups in 2023 alone. As early movers and now the industry standard (c.80% market share), Nvidia's A100 and H100 chips are not only the most efficient option, but also crucial for large scale AI training and inference (chips are the biggest cost head for LLM companies). With unprecedented interest in AI (ChatGPT acquired 1M users in 5 days of launch, OpenAI reported $3.4B ARR in Apr’24, AI agents are widely recognised as the tech of the future), Nvidia stands to capture a large portion of the $250-$500B data centre conversion market over the next 5 years​. This is proven by the $22.6B in data centre revenue Nvidia brought in during Q2’24, up 23% from the previous quarter and up 427% from a year ago 🥇Competitive positioning: While Intel, AMD, Arm have been active, the only credible competitor, AMD, has a relatively smaller market share in the high-performance AI GPU segment, and Nvidia’s consistent technological advancements keep it further ahead, e.g., Nvidia’s Rubin which will succeed Blackwell in 2026 is orders of magnitude better than anything on-market. 🤝Moats via ecosystem building: Collaborations with hyperscaler cloud providers like AWS and Microsoft Azure have expanded Nvidia’s reach. Moreover, acquisitions like Mellanox have strengthened its position in high-performance networking, critical to AI infrastructure. Coreweave, a prominent cloud services partner / distribution channel for chip capacity is an NVIDIA investee and is built almost entirely on Nvidia chips. Finally, they continue investing in networking, software tools, data centre solutions but also purpose-built AI PC and smartphones to create an ecosystem that not only locks in customers and enhances switching costs, but pivots more of their revenue to more predictable recurring contracts. 📈 Financial underpinnings: Nvidia’s expected FCF for next 2 years stands at $100B, providing huge ammo for R&D, strategic investments and M&A (only $30 to $35B earmarked for buybacks) and in turn, growth. Even so, their shares trade at a P/E ratio of c.20x  NTM earnings, far below historical median of 35x+, indicating the market expects them to grow in the medium term. KEEN TO HEAR THOUGHTS! #Nvidia #AI #Compute #Chips #Tech #Growth

10

1 Comment

https://lnkd.in/dSQGZg_T "They warned that such AI systems could perpetuate the racism and sexism found in the huge troves of written data they've been trained on. "The problem with that kind of misinformation is that we're swimming in it," Bender said. "And so people are likely to get their biases confirmed. And it's harder to spot misinformation when it's confirming your biases." Another concern was a deeper one — that ceding information retrieval to chatbots was degrading the serendipity of human search for knowledge, literacy about what we see online, and the value of connecting in online forums with other people who are going through the same thing. Those forums and other websites count on Google sending people to them, but Google's new AI overviews threaten to disrupt the flow of money-making internet traffic. Google's rivals have also been closely following the reaction. The search giant has faced pressure for more than a year to deliver more AI features as it competes with ChatGPT-maker OpenAI and upstarts such as Perplexity AI, which aspires to take on Google with its own AI question-and-answer app."

2

1 Comment

DeepSeek V3 - new open model that is better than Claude Sonnet at the Aider Polyglot coding benchmark and - perhaps most importantly - has 10x cheaper inference costs. The massive price drop means you can experiment with having a Aider build you a feature for $2 instead of $20. $20 is a little too pricey for rapid experimentation in your development workflow. Unsurprisingly, Deep Seek's site is down. Other Model Inference providers like Hyperbolic seem to have also taken Deep Seek V3 down for maintenance https://lnkd.in/gtn8Hb36

9

Interesting experiment for web3 communities. Step 1: Pull 100 random people Step 2: Ask them to explain the project Step 3: Validate if they can I'll bet most community members can't even explain the project and are likely sitting there for airdrops, giveaways and more. A byproduct of hype-based marketing that leads to big numbers that don't convert to value. 🚀 Follow Tony Drummond to learn all about web3 ---- Why listen to me? – 3,500+ hours experience running projects in web3 – Scaled my project to 75,000 holders in 3 weeks – Made every mistake in the book – Learned from near all of them – Follow me to avoid them

45

30 Comments

New spin-off from the SRI / NSG labs at ETH Zürich, solving the network observability problem with GenAI & mathematical models, based on decades of research work on automated reasoning for network verification and synthesis. Follow NetFabric.ai to track where this goes :)

34

🌐✨Exciting times in the AI world! French startup Kyutai has unleashed Moshi, an open-source, real-time multimodal language model. Imagine a voice assistant that listens, speaks, and understands emotions—all with an impressive 160ms latency. 🎙️ What is even more mind blowing? Kyutai’s lean team of 8 engineers pulled this off in just 4 months. French engineering brilliance at its finest. But here is the real magic anyone with determination can achieve the seemingly impossible. Whether you are a seasoned developer or just starting out, remember that innovation knows no boundaries. Stay tuned for the code, model, and research they are about to democratize voice AI. Let’s embrace the future! 🤖🌟 #AI #VoiceTech #Innovation

1

35 Days into YC: 3 Real Startup Lessons We Didn't Expect 🪄 #1 Master the Art of the Impromptu Pitch - Keep it magical: Your demo should always tell a story — keep the focal point your users, and make sure they experience the “WOW” moment of your trick. 🎯 #2 Activation Trumps Acquisition: The Real Metric That Matters - At Spur (YC S24) we measure activation by measuring the frequency with our users are creating and running tests on the platform. - Signing up for a demo, a pilot, or even onboarding does *not* signal activation The question that really matters: are your users actually using your product and finding value from it? ✨#3 Make things ridiculously easy for your customers - Set up easy channels of communication with your customers, and be proactive about using them. - Handoff doesn't cut it, your customers should hit the ground running when you onboard them — even if it means you are doing the work to set them up P.S. 👀 Keep an eye out! Exciting things to come from us early next week #ycombinator #AI #Testing

233

11 Comments