Brad Terrell

As LLMs start trending towards commoditization, a pressing question becomes important – how can companies in the recruiting space use AI to maintain a competitive edge? Although we are seeing some strong indications that AI and specifically LLMs trend towards commoditization across all industries, there are specific scenarios where technological differentiation is possible. 🚩 Proprietary data In a world where everyone can access the same models with a few clicks, the value lies in proprietary data and not prompt engineering. Proprietary data has strategic applications in generative AI. It can be used for fine-tuning, retrieval-augmented generation (RAG), and additional prompt input. Taking this further to the recruiting industry, the value will come from unique data and from owning quality data that is diverse, relevant, and collected in an ethical, AI-policy & privacy-compliant way. Generally, I wouldn’t say I like writing about AI ethics, but in this case, I think it is worth talking about. The AI policies and the dynamics around their enforcement will only continue evolving. In the context of HR tech, where sensitive personal information is involved, how data is processed, interpreted, and utilized might be as crucial as owning exclusive data sets. Preparing for a future landscape where the legal integrity of data will be challenged a lot more often will help companies adapt to legal changes more swiftly. Are there other differentiation factors? 💡 Well, there is UX. Objectively, AI-powered products in all industries might require different UX or even devices from what we are used to today. There is undoubtedly some edge here, but considering the data collection path, it is much easier first. In conclusion, HR Tech companies looking to build a competitive edge using generative AI can leverage proprietary data ethically, enhance user experience, and stay ahead of regulatory changes.

5

2 Comments

We're launching the inaugural edition of Beyond Benchmarks, a deep dive into the metrics and trends observed across the early-stage enterprise cloud market. This is the first look at some of the business data around GenAI in SaaS. Some initial observations: 1) 60% of SaaS companies have already integrated GenAI into their offerings (60% of these as part of their current product and 40% as a new product), and an additional 20% are planning to do so this year. Almost half aren't monetizing these features. Those that are are primarily experimenting with both usage-based and flat fee pricing models. 2) Most companies are utilizing OpenAI as their primary LLM, but many are experimenting with multiple models - and we are noticing an early trend towards intelligently routing GenAI inference requests to different models based on cost, performance, and security. 3) Though we are in the initial innings of GenAI, we are seeing some positive signals - companies that implemented GenAI had 7% higher NDR than those that did not. Still early days, but we'll be doing this benchmarking report regularly to track and share how things progress in this new era. Link to report in comments.

104

6 Comments

Interesting though unsurprising finding from Bain survey on GenAI implementations thus far at large companies: primarily focused on incremental optimizations, which makes sense! Even what Bain characterizes as "growth" is likely dominated by incremental GenAI additions to current products, rather than truly "transformative" applications.

2

Starting this Monday, Oct 7, on the first Monday of each month, CryptoMondays NYC will focus on the intersection of crypto and Wall Street (i.e. institutional participation in crypto assets). Please join us Monday as we kick off this initiative with a Fireside Chat with Zach Pandl, Head of Research at Grayscale Investments. Grayscale's iconic Bitcoin Trust, started in 2013, peaked at over $40B in AUM in 2021. Today, Grayscale offers more than 20 crypto trusts (e.g. ETH, Bittensor, & Solana) totaling more than $20B in AUM. Prior to joining Grayscale, Zach was the Senior Economist/Macro Strategist at Goldman Sachs. I'm looking forward to seeing lots of my Wall Street friends who range from curious about crypto, to obsessed. Anthony Scaramucci, Alexander Golding, Anoop Kansupada, Anthony Pompliano Catherine Wood Yassine Elmandjra Dave Balter Hadley S. Thomas (Tom) Lee Matthew Walsh Miko Matsumura Pete Najarian Mo Shaikh Anna Bertha Evan F. Seth Ginns Jalak Jobanputra James Parillo Mark Goldwasser Keli Callaghan Mike Chan Nolan Ford Max Shapiro, CFA, CPA Sunny Parikh Ashim Mehra Wendi Dritz Carver Tyler Carter Michael Gold Mike McGlone Matt Wright Marcy Simon Casey Taylor Brad Feinstein Jonathan Steinberg Gil Rosen Rick Bradt Kenneth Kruger Carl Vogel #bitcoin @eth #btc #crypto #cryptocurrency #web3 #AI https://bit.ly/CM1007ZP

64

5 Comments

Interesting answer from AlphaSense CEO Jaakko (Jack) Kokko on why they acquired Tegus - all about access to proprietary data, in this case the transcript library on private companies. You recently acquired a company called Tegus. Was this about having more proprietary data to feed into your models? First and foremost, it was about providing our customers with the best possible intelligence on the market, especially on private companies. A couple of years ago, we acquired a company called Stream, the No. 2 player in what you might think of as a library of interview transcripts between investors and experts on a company — their former executives, competitors, customers, partners. It allowed us to produce information that you can’t access anywhere else, and our customers loved it. Tegus is the market leader in that sector and was the natural next step for us. They have the world’s best library of private company content, and we immediately wrapped our search capabilities, our generative AI, our monitoring capabilities around their content and made it much more valuable. We now have the broadest coverage of insights on both public and private companies, allowing our AI to provide even better answers to our customers.

161

8 Comments

I'm thrilled to announce the launch of 𝐈𝐑𝐈𝐒 𝐅𝐨𝐮𝐧𝐝𝐫𝐲 (𝘩𝘵𝘵𝘱𝘴://𝘸𝘸𝘸.𝘴𝘺𝘮𝘱𝘩𝘰𝘯𝘺𝘢𝘪.𝘤𝘰𝘮/𝘪𝘯𝘥𝘶𝘴𝘵𝘳𝘪𝘢𝘭/𝘪𝘳𝘪𝘴-𝘧𝘰𝘶𝘯𝘥𝘳𝘺/), a cutting-edge DataOps platform built on Generative and Predictive AI, that we've been passionately working on for the past several months. This platform includes built-in pre-trained models and recently launched Industrial LLM in partnership with Microsoft. I want to extend a heartfelt thank you to all the teams who have worked tirelessly to make this possible. The initial response has been phenomenal, and it's truly a testament to the hard work and dedication of everyone involved. 𝐈𝐑𝐈𝐒 𝐅𝐨𝐮𝐧𝐝𝐫𝐲 is an advanced DataOps platform engineered to empower organizations by transforming raw data into actionable insights. It integrates effortlessly with existing enterprise, IT, and OT systems to collect and give context to data from diverse sources, utilizing predictive and generative AI technology to make it accessible and actionable. The platform offers real-time KPI monitoring, provides detailed, contextual insights, and features predictive functionalities powered by advanced AI. It also enriches data through an integrated Industrial Knowledge Graph and simplifies organizational access with a Unified Namespace. With hundreds of connectors, IRIS Foundry streamlines data integration into its built-in data lake, reduces the complexity of network configurations, and minimizes the need for multiple data connections. This promotes effective data modeling, contextualization, and governance, ultimately boosting decision-making and operational efficiency. https://lnkd.in/g85_KqHG

94

7 Comments

🚀 Generative AI Predictions for 2025 ✨ 🔺 Prediction: Prompt Data Becomes the New Gold Standard 🔹 In 2025, prompt data—the inputs users provide while interacting with AI tools like ChatGPT, Gemini, and Claude—will emerge as a key commercial asset.  Linked to identity, this data offers rich insights into user intentions, preferences, and purchasing plans, making it highly valuable for marketers and tech companies. 🔺 Prediction: Specialised Small Language Models Will Lead in 2025 🔹 By 2025, small language models (SLMs) will outperform larger models in cost, speed, privacy, and adaptability. SLMs have significantly lower inference costs due to reduced computational requirements, making them ideal for businesses prioritizing efficiency and scalability. SLMs are poised to become the backbone of enterprise and domain-specific AI products. 🔺 Prediction: Web3 and GenAI Will Enable Tokenized Human Intelligence by 2025 🔹 By 2025, Web3 and generative AI will converge to create tokenized human intelligence as a service. Blockchain and tokenization will allow individuals to contribute valuable data and insights in exchange for digital assets, enabling decentralized crowdsourcing to improve AI models. This will lead to more personalized AI outputs and equitable data management. 🔺 Prediction: Distributed Computing for Inference Will Revolutionise AI in 2025 🔹 In 2025, distributed computing will become a key method for handling the growing demands of AI inference. New orchestration layers will enable the seamless connection of thousands of individual CPUs and GPUs, creating powerful distributed networks capable of executing complex inference tasks. The ability to harness large-scale distributed systems for inference will democratize access to high-performance AI, making it more accessible and cost-effective for enterprises across industries. 🔺 Prediction: Cybersecurity and Governance Challenges Rise with GenAI 🔹 As generative AI evolves, cybersecurity and governance concerns will escalate. Deepfakes and advanced attacks like undetectable phishing will grow, while AI will also improve threat detection. Businesses and governments will prioritize AI governance and security, with regulations like the EU AI Act pushing for stronger data privacy and accountability measures to manage these growing risks. 🔺 Prediction: GenAI Will Shift Focus from Cost Savings to Revenue Generation 🔹 While GenAI has been widely used for cost savings, its focus will increasingly shift to revenue generation. Use cases like AI-driven sales development representatives (SDRs) and enrichment tools have created sales pipelines but haven't fully translated into closed deals. In 2025, businesses will move toward leveraging GenAI to enhance customer value, increase conversions, and unlock new revenue streams, marking a shift from operational efficiency to driving growth. What do you think about the future of Gen AI? Lets discuss! 🌟💬 #genai #slm #prompts #FutureofWork

5

1 Comment

Model Explainability means different outcomes for different users. We can't have one size fits all when explaining model prediction. At the same time, unless the explainability is true to the model, it has credibility issues! All regulators expect to have explainable ML models, which is not standardised! How do you scale ML explainability in high-risk and high-compliance use cases? Continuing the article in Forbes, I've written about various XAI outcomes and how to approach setting up such explainability. #AryaXAI #ExplainableAI #EnterpriseAI #AIRegulations https://lnkd.in/gp7YgzB3

13

1 Comment

I recently had the opportunity to discuss OpenAI’s upcoming model, ‘Strawberry,’ with PYMNTS. This new advancement is not just about the technology—it’s about how AI is reshaping industries. From optimizing supply chains to forecasting market trends with precision, AI is creating personalized customer experiences at scale like never before. The future is here, and it's AI-powered! 💡 Read more here: https://lnkd.in/eZtR_y-s #AI #Innovation #BusinessTransformation #OpenAI #StrawberryModel #TechTrends #FutureOfBusiness #ArtificialIntelligence

66

1 Comment

How should OpenAI price Strawberry or o1 as it is now known. Flash poll results. Total of 74 responses from Friday Sept. 13. Interested in how generative AI applications are being monetized? Join Kyle Poyar from OpenView and Growth Unhinged, Michael Mansard from Zuora and Steven Forth from Ibbaka for an engaging roundtable on Sept. 19. Please register here. https://lnkd.in/gZNZqS6u My own thoughts is that this depends on strategic intent. Strawberry is a very different application than vanilla ChatGPT and it will provide value in new ways. It should not be tied to ChatGPT pricing. I would position it as a separate offer. However, part of the intent of Strawberry seems to be to provide training data for future reasoning models. In that case the more people that use o1 the better and pricing should be neutral. #Strawberry #Openai #generativeAI #genAI #pricing #monetization #o1

7

3 Comments

As we continue to push the capabilities of Vizerto by integrating agentic workflows using Generative AI techniques. Here are some of the recent learnings and insights: 1. User Trust and Transparency Challenge: Business users are less forgiving of mistakes made by AI compared to human errors. Example: An AI-generated answer, sourced from a Slack thread where an expert provided the solution, was deemed incorrect by users. If the same answer came directly from the expert, it might not have been questioned due to their established credibility. Insight: Users tend to scrutinize AI outputs more critically because machines lack the inherent trust that human experts have earned. Solution: We implemented full traceability for all AI-generated answers. Users can now see the exact sources—like specific Slack threads—that the AI used to generate its responses. 2. Access to Essential Information Challenge: Don't assume but AI agents need the same access to information as human users to perform tasks effectively. Example: The AI agent was unable to proceed with tasks involving specific opportunity and order link that were only accessible to specific owners within the organization. Insight: Without appropriate permissions and access to necessary applications or documents, the AI cannot replicate the tasks a human would perform.  Solution: Ensure that AI Agent access has the same access levels as human users to perform a given task. Don't just assume API calling would do the trick. 3. Latency and Learning Efficiency Challenge: Agentic workflows involve planning, reflection, tool usage, and iterations, which can increase response times compared to instantaneous human actions. Example: A task that takes a human a few seconds—like recognizing similarities between procedures and skipping redundant steps—might take the AI longer because it processes each step individually without inherent shortcuts. Insight: Unlike humans who intuitively streamline tasks based on experience, AI may not optimize them unless it retains extensive contextual information. Solution: optimizing our AI's performance by enhancing its context retention capabilities and teaching it to recognize and adapt to repetitive patterns. #AI #GenerativeAI #WorkflowAutomation #Vizerto

37

1 Comment

In an astonishing turn of events, the cost of intelligence is plummeting to nearly zero and becoming omnipresent. GPT-4o mini's latest pricing announcement is just the tip of the iceberg. Recent statements from multiple vendors reveal that competent large language models are not only becoming cheaper but also faster, more accessible, and remarkably better. One jaw-dropping statistic accompanying the GPT-4o mini news is the 99% drop in inference costs since 2022, during which time the models have grown significantly more sophisticated. The open-source community is buzzing with activity, churning out competent models weekly, many of which can run on everyday consumer hardware. New small language models (SLMs) are emerging that can operate efficiently on laptops and even smartphones. Never before have we witnessed such a dramatic and swift transformation in the tech landscape. This is just the beginning, with new capabilities and efficiencies set to emerge over the next 12 to 24 months. Stay tuned for more groundbreaking advancements.

7

1 Comment

It seems we have got a very bright future coming super soon with AI. That'll bring in creativity, perspectives, opinions that too with speed for accurate decision making. Must watch this quick demo by Allie K. Miller and Reid Hoffman 👇 Looking forward to exploring more and more and more! #AI #gpt #openai #linkedin

12

Thanks to #genai360 for the shoutout here for Consilience AI and our AlphaIQ product. Davit Buniatyan Joseph Byrum, Ph.D., MBA

10

NVIDIA tops the Axios Harris Poll 100, which ranks company reputations. FYI, this poll is done using this three-step process: 1. January 2024 Survey: 6,273 Americans identified the 100 most visible companies. 2. March 2024 Survey: 16,500 Americans rated these companies on nine reputation dimensions, creating a Reputational Quotient (RQ®) score. 3. May 2024 Surveys: Two waves of surveys provided additional context on brands and politics.

6

1 Comment

The Elephant in the Room (of AI): Memory Match Up! Can you remember where you left your water bottle last week? Elephants, remember! These brainiacs with billions of neurons can remember watering holes from years ago. Talk about an impressive memory! AI is no elephant (yet!), but GenAI applications built with Karini AI are mastering some fascinating techniques. We don't rely on biological memory but on a unique 'semantic cache' that acts as GenAI's memory jog. It stores recently used information, so when a question resurfaces, it's like a light bulb moment! I remember that!' Just like an elephant recalls a familiar watering hole. Here's the twist: the elephant's memory constantly learns and adapts based on experiences and emotions. Our semantic cache, on the other hand, relies on data and algorithms. It's clever but not quite as "elephantastic." So, who wins the memory battle? It's a tie! Elephants are champions at the long game, while GenAI shines with lightning-fast retrieval. Together, they demonstrate the incredible potential of both nature and technology, inspiring us with the possibilities of AI. And guess what? Karini AI's managed semantic cache takes things a step further. By efficiently managing and optimizing the storage and retrieval of information, GenAI applications run even faster and more intelligently, saving you money. Pretty cool. 😎 Ready to dive deeper and unlock the power of AI memory? You can find out more about what we're doing at Karini AI. The possibilities are endless. ➡ https://lnkd.in/eSpifX5a

4

Incisive analysis by Sequoia partner David Cahn on the missing $600 billion in AI revenue... Basically, given what companies are spending on Nvidia's chips and infrastructure build out, a certain amount of resulting revenue is implied. There is a $600 Billion gap. This week's The Economist arrived at a similar conclusion. The ROI will likely be realized after 2032! Links below. #ai #chatgpt #nvidia #venturecapital #startups

8

1 Comment

Service-as-software is a $4.6T opportunity. But no one knows how to actually grab that opportunity yet. Enterprise software has gone from assisting human workflows to now, LLMs giving software the ability to actually perform parts of the workflow. Instead of relying on timely and sometimes inaccurate human inputs of structured data, today's systems actively capture and process both structured and unstructured data. The service-as-software shift was huge—but we're on the edge of one that's even bigger. My Foundation Partners Joanne Chen and Jaya Gupta have now outlined how to put the service-as-software paradigm shift into action: through a System of Agents, which doesn’t just streamline a workflow but actually completes whole tasks, becoming the worker. A System of Agents mirrors a human team, with agents taking on collaborative, specialized tasks and continuously learning from each other, just as human teams do. Advice for founders building with Systems of Agents: → Position at the data source, own the path. Traditional software waits for humans to input data, whereas Systems of Agents capture it at the source. When an agent owns the interface where data is born, the system can then orchestrate every downstream action. → Move beyond software budgets, tap workforce spend. When a System of Agents completes a whole job task, it should be categorized as a personnel cost, not a software expense. → Use AI’s 24/7 nature and scalability to create new workstreams. AI agents working in areas like healthcare can operate 24/7, providing reliable support and an around-the-clock presence. → Stay ahead of pricing and business model shifts. Service-as-software shifts the core business model from seat-based pricing to outcome-based pricing; value is placed on the tangible improvements that AI delivers. In the past 6 months, we’ve heard from hundreds of startups working in this space. After those conversations, we’re announcing the System of Agents 50, which includes the companies you see here. 👇 Full blog in the comments. 

736

65 Comments

From Azeem Azhar on Eric Schmidt’s recent talk at Stanford: “Agent-based systems, which autonomously execute multi-step tasks and adapt to environmental feedback, are poised to unlock significant economic value next. Building on this foundation, text-to-action capabilities will emerge, translating natural language into specific, working programs. Imagine your own personal programmer at your fingertips, ready to create whatever you envision. A dating app that matches people based on their Netflix viewing history and Spotify playlists? A prototype within minutes. These forces will converge. While he doesn’t say it explicitly, it seems that Eric expects the next two years to be faster and more turbulent than the previous two.” https://lnkd.in/gxWEy5Pf

8