Bob Bauer

🚀 Accelerating Insights with #VentureIQ 🚀 One of the most exciting things about #VentureIQ, as Neal Ghosh and 9point8 Collective say, is how quickly it distills complex concepts across diverse fields—whether it’s #AI, #deeptech, or patent submissions. By leveraging contextual learning, VentureIQ rapidly uncovers key business insights for investors and founders alike. The ability to move from dense technical material to actionable commercial strategies in minutes is a game-changer for venture-building. Curious how it works? DM me and let’s chat! Check out the original post for a deeper dive ⬇️

2

Batch processing for LLM inference activity: "One sign of that is the rise of batch processing. Today, popular consumer AI products like ChatGPT and Perplexity provide users with near instantaneous responses, otherwise known as real-time inference. However, businesses don't always need immediate responses and are often willing to wait hours, days, or even weeks for responses, as long as they don't have to shell out as much money. Founders of cloud and inference providers have told me there's growing pressure from business customers for this kind of flexibility, otherwise known as batch processing." -- Nathaniel Whittemore on "The AI Daily Brief" (with much deserved shout-out to his inspo from Stephanie Palazzolo at The Information) I fall into this bucket as we have a lot of backfilling to do as we're building out the underlying data in our service, and some of that could just as well be done in the wee hours of the night. The new stuff flowing in I want processed ASAFP, but these backfills—I'd mos def take more cost-efficient delayed batch processing. There's also use cases like needing (near) real-time inference for the first inference pass for timeliness / contemporaneousness value, but then a second pass slow-time inference for wider classification exercises over an ever-growing data corpus. I could see us mixing Sonnet & Haiku models in both cases. Just a heads up to Anthropic Mike Krieger as you contemplate this, along with Amazon Web Services (AWS) Jeff Barr Mark Roy Swami Sivasubramanian on the Bedrock side of things...and I don't mean the "batch" mode that doesn't seem to have any real value other than helping AWS manage its own rate limit challenges. (DeepCast is using Bedrock.)

11

2 Comments

Quantization is a game changer for deploying large language models like Mistral 7B on consumer hardware. This course offers invaluable techniques to harness powerful AI models even with limited resources.

2

Three takeaways here: 1. #VentureIQ is insanely good at analyzing dense #RandD and #IP reports and detailing the potential applications of the underlying IP. That said, the reports still skew towards a technical audience... 2. #NotebookLM is insanely good at taking technical writing and producing "podcasts" that a 10 year-old can understand. If you don't believe me, just listen to the first 30 seconds: https://lnkd.in/ep-FNc_z Together #VentureIQ and #NotebookLM can take complex material seemingly written in a foreign language, distill what it does and why it matters, and then explain it to a 10 year-old. And it only takes 5 minutes. Pretty cool, right? This brings me to item #3.... Neal Ghosh, is crazy smart (he ran this on his own #ML patent) and has been working on #AI since before it was cool. Everyone should have a Neal on their team.

6

Congrats VERSES (OTCQB:VRSSF) for: Introducing "an efficient alternative to deep learning, reinforcement learning and generative AI".... that "achieved 99.8% accuracy on a subset of the MNIST digit recognition task, a common benchmark in machine learning, using only 10,000 training images (90% less data)" Hey Gary Marcus did you see this? #ai #activeinference #spatialweb

1

Arvind Narayanan's latest talk delves into the complexities of evaluating large language models (LLMs). A must-view for those keen on understanding the challenges and nuances of LLM evaluation in AI development. Catch the full talk here: https://buff.ly/3ZG68Rq #AI #LLM #Evaluation #TechTalks #ArvindNarayanan #AIResearch

5

Regulatory oversight can provide short-term stability and protection, but if not carefully planned, it can often stifle innovation. The SEC's regulation of #ICOs in 2018 is a prime example. The increased regulatory burden and uncertainty led to many ICOs excluding US investors and positioning themselves as foreign offerings to avoid SEC jurisdiction. This shift limited access to capital for US-based early-stage startups and created new challenges for companies seeking to raise funds through ICOs. Most importantly, the lack of clear guidance made startups hesitant to explore new financing models, fearing regulatory repercussions. It's ironic that a #gambler can lose their house in minutes in #Vegas without the same level of scrutiny and regulation on #casinos! And kudos to Office of California Governor Gavin Newsom for not getting too excited or bowing to pressure! If the government wants to be part of innovation, they should offer grants and non-dilutive funds to startups, just like #Europe and #Canada do with - SR&ED Tax Incentive Program - AgriScience - NRC-IRAP - EU Grants for Digital Health and Technology Startups - IIA Investment Programs and many more! This way, they can keep an eye on startups and disruptive solutions as they submit their progress each year to qualify for more funding. Let's support our startups the right way! #InnovationBeforeRegulation #AI PBS KQED NPR BBC CNN

9

3 Comments

My epic co-founder Tim Cheung did an excellent job (as always) in his latest post for Factory AI. The article is about how manufacturers can leverage the goldmine they’re sitting on—their data. First, he clears some of the fog on today’s hot topics: AI, ML, Deep Learning, and Gen AI. Second, he shares six relevant use cases, each accompanied by a case studies for each. Check out the link to the post in the comments. Give it a read and let us know your thoughts as always. #Manufacturing #FoodAndBeverage

16

3 Comments

Lately I've been curious to learn more about how AI is impacting biology and drug development. With all of the hype around Generative AI for text, there seems to be a growing discussion around Causal AI to identify new proteins and predict treatment outcomes. I thought I'd share a few helpful things I've read in the area: This article from Century of Bio offers a great explanation of how causal AI is used for protein discovery. It also gives a really approachable breakdown of how this kind of AI works. https://lnkd.in/g3U3fA4D This article from Nature Medicine goes into detail around the benefits of causal AI for predicting medical outcomes and how it can lead to better individual patient predictions: https://lnkd.in/ggnsvD4z This paper looks at the early success of AI developed drugs in the place where it matters most: clinical trials. It's interesting to see that the early pipeline of drugs designed using AI are in line with historical averages for classically developed drugs (around 40%) https://lnkd.in/g3Y6Jf5S I'll be keeping a close eye on this space! #ML #AI #CausalAI #Biology

20

📢 Exciting news for #healthcare and medical AI. Advanced reasoning and specialized, timely updated medical knowledge are crucial for AI models to assist in clinician, researcher, and patient workflows. In this context #LLM are challenged by the speed at which medical knowledge changes and expand: from new #drugs, #therapies, #regulations, etc… the body of information that is needed to make AI valuable requires unusually complex #RAG architecture. #google introduction of Med-Gemini, a new family of next-generation models fine-tuned for the medical domain is a major step forward. Gemini models have demonstrated a significant leap forward in multimodal (textual, imagery, video, tabular, etc…) and long-context reasoning, making them a perfect fit for the complex and extensive data found in #electronichealthrecords, #ehr. With the introduction of Med-Gemini, this jumps to the next level by fine-tuning on de-identified medical data.🚀 #AIinHealthcare #MedicalAI #GeminiModels

10

We are thrilled to announce the release of OpenBioLLM-Llama3-70B & 8B from Saama AI Research Labs. These groundbreaking models outperform Openai’s GPT-4, Google’s Gemini, Meditron-70B, Google’s Med-PaLM-1, and Google’s  Med-PaLM-2 in the biomedical domain, setting a new state-of-the-art for models of their size. Both models underwent a rigorous two-phase fine-tuning process using the LLama-3 70B & 8B models as the base and leveraging Direct Preference Optimization (DPO) for optimal performance. This release is just the beginning! In the coming months, we'll introduce expanded medical domain coverage, longer context windows, better benchmarks, and multimodal capabilities. Stay tuned for our upcoming official paper. This work was led by Ankit Pal Bhaskar Sambasivan Lisa Moneymaker Ari Srinivasan Rajeev Dadia Pravin Kumar Prithiviraj S Kamal Raj Kanakarajan Tamil Arasan Ph.D. Murugesan Vadivel Selvakumar Murugan Mahesh Narasimhan Basant Shanker Sas Maheswaran Panikos Christofi Rajasekar Nonburaj (𝑅𝒥) Navin Kumaran Vikas Kishore Geetha Eswaran Ojasvi Harsola Angu S K. Ramanathan Srinivasan Model download link is in the first comment.

839

106 Comments

A bill pending in the California Senate that would essentially make many start-ups, open source projects as well as some research illegal in California and force it to move to other states. We need to stop this nonsense. Usually I don't post about politics as issues are often nuanced, complex and hard to fully address in a short post. This is an exception. The bill has zero benefit for anyone, except maybe very large tech companies that can use it to crush stat-ups. And it doesn't mitigate any potential downside of AI. However it would kill many open source efforts. It would stop some research. And it would force start-ups to do their model development in other states (or other countries). If you can, please support the effort to stop this. https://stopsb1047.com/

55

5 Comments

The MUSE Benchmark Paper shows that current unlearning techniques used to remove private and copyrighted content leads to model degradation. This reinforces the discussion about the need for high quality data with provenance tracking at very beginning of the model training process. Read more about it here: https://lnkd.in/dpipZ4cf

18

Akhil Rachamadugu - this is going to be a great session! Who is going to World Forum next week? Let’s connect! 🚀 USDM will be there, sharing how we extend ServiceNow with ProcessX to support regulated use cases across clinical, lab, manufacturing, post-market, and quality operations. We are enabling AI Agents in ProcessX for further automation. With ServiceNow + ProcessX + AI, intelligent agents can handle repetitive tasks and business can focus on driving greater value! Let’s explore how AI-powered workflows can transform your operations and elevate business outcomes. Looking forward to connecting!

7

1 Comment

From where I sit in the SF Bay Area, cradle of innovation for at least the last 30 years, I'm a true believer in the promise of the #AI revolution. That said, I find it fascinating to consider views on the opposing side. I'm sharing below two opinion pieces from the last two weeks suggesting that the AI revolution is fundamentally overhyped and a bubble ripe for bursting, and that AI tools are actually not useful or transformative. Contrast this with venture capitalist Kai-Fu Lee's prognostication that 50% of white collar jobs will be displaced by #AI within three years, and that OpenAI and the other cutting edge movers in the space are headed to a trillion+ dollar valuation. Which side will win out? Where will we be by 2030 in the AI race? I'd love to hear your thoughts in the comments.

7

3 Comments

When you’re bootstrapping a demo, a large, closed-source AI inference platform might be perfect for your needs. But what happens when you hone in on a specific use case? Fireworks AI is emerging as a market leader that gives companies the ease, latency, price, quality and ownership they need as they reach product-market fit—and access to compound AI systems once available only to a select few. We at Sequoia Capital are excited to double-down on our partnership with Lin Qiao Dmytro Dzhulgakov Benny Yufei Chen Chenyu Zhao Dmytro Ivchenko James Reed Pawel Garbacki and lead their Series B.

170

12 Comments