Nicholas Clarke

Excellent report from General Dynamics Information Technology John P. Curran regarding adoption and trends of AI in the Govt. I used Meta to generate summary below and included link to original article. Main AI use cases are chat and creating/updating computer code. Note also the 23% failure rate of AI projects due to user expectation misalignment. This can be overcome with thorough protoyping and establishing use case metrics. - Successful AI project completion timelines are narrowing, with an average of 14 months from pilot to production. - Future demand for AI projects is building around chat and coding-related functions. - 54% of surveyed AI experts said projects currently under development involve improving computer code. - 59% of projects involve "research and understanding", which may relate to creating chat functions. - 23% of AI projects fail due to misalignment between AI solutions and user expectations. - Partnership efforts are required for successful AI projects. Data diversity is crucial, and incorporating diverse data sources can lead to better AI outcomes. Trends: - AI adoption is increasing across the Federal government, with a growing focus on chat and coding-related functions. - Agencies are leveraging cloud-native services, tooling, and other capabilities to accelerate AI project development. - There is a growing interest in translating older code to newer languages, such as Python or cloud-based analytic tools. - Domain-specific chat functionality is becoming a priority, with many agencies planning to implement chat services in the next two years. Statistics: - 325 AI experts surveyed across the Federal government - 14 months: average time from pilot to production for AI projects - 54%: percentage of projects currently under development that involve improving computer code - 59%: percentage of projects that involve "research and understanding" - 23%: percentage of AI projects that fail due to misalignment between AI solutions and user expectations. Bill Casola John Grim Leslie Anne McIndoe Kim Giffin Michael S. Freeman III Tobias Quaranta https://lnkd.in/eqvmAVFG

3

Microsoft's AI model, Aurora, quickly predicts levels of six major air pollutants worldwide, including carbon monoxide and particulate matter. Read more in this article from Nature. #msftadvocate #ai #nature #aurora https://lnkd.in/g-AppzdB

2

1 Comment

This is a long piece about the many sessions on AI at last week's #GartnerSYM, including advice on picking the right kind of AI, scaling systems, and governance. Some good practical advice. https://lnkd.in/erpnXZ9v

10

2 Comments

Another demonstration of ScienceCast.org #AI enhancements of arXiv postings - free, simple interface https://lnkd.in/e_gXsF_9

3

1 Comment

AI is navigating a very tricky transition from "novel" to "useful" at the moment, and while the wild creative swings and the (now-maligned) wearables are capturing the attention, there's a deeper discussion happening about what it means to get generative models ready for reliable, robust, enterprise-grade deployment. More in our breakdown of this week's top #fashiontechnology news: interli.net/4aYFUhn

38

4 Comments

Bit by bit, we're slowly reinventing the semantic web. The article here explores an idea that I've been thinking about as well. There are two key advantages to integrating knowledge graphs via RAG vs. trying to work JUST from LLMs - the fist is that you have the ability to name entities and taxonomies, ensuring topical synchronization via taxonomies. However, a second aspect is that you can teach an LLM to utilize a specific data structure by passing it a schema. However, there's a corollary to that. If you can teach an LLM how to build schemas, then the LLM can use this to design a schema for its own purposes to optimize reasoning. Now, I'm going to make a contrafactual argument here to others (people I respect) to say that I think an LLM is perfectly capable of reasoning, because we don't understand reasoning. An LLM is incapable (and even here I'd quibble) to utilize intent, but if it is specifically informed to create a "castle in the sky" graph that allows it to perform reasoning tasks, it should be able to traverse that graph with appropriate constraints, because I've used that technique myself when using LLMs to design schemas. Put another way, in order for an LLM to reason, it needs to become an ontologist. https://lnkd.in/gW6D9TeW

44

6 Comments

Can AI Evaluate AI Better Than Humans? The quest to make LLM-as-a-Judge a reliable, cost-effective alternative to human evaluation for text quality is evolving fast. But there's still a gap—human evaluators bring a nuance that current AI lacks. Enter RevisEval, a novel approach designed to narrow that gap by using response-adapted references. Here’s the big idea: Instead of using static references, RevisEval uses the text revision capabilities of large language models (LLMs) to dynamically revise responses and create highly relevant "adapted" references. This means that the reference itself changes to better reflect the response being evaluated, making the evaluation process more accurate. The results are impressive—RevisEval outperforms traditional evaluation methods, boosts metrics like BLEU and BERTScore, and even rivals human evaluation in certain scenarios. This could be a game-changer for AI evaluation—making assessments more dynamic, relevant, and accurate, while still being highly scalable. Interested in how this works and why it matters? Check out the details here: https://lnkd.in/gHhPUKSq #AI #MachineLearning #TextGeneration #LLMasJudge #RevisEval #NLP #ArtificialIntelligence #Innovation #LLM #NaturalLanguageProcessing #AIResearch #TechInnovation

4

Is generative AI a friend or foe in innovation? In 2023, the U.S. Patent and Trademark Office made a bold move. They banned the use of generative AI tools like ChatGPT for their staff. Why? They fear that reliance on AI could compromise the integrity of the patent process. Imagine a world where creativity is stifled because we're too dependent on technology. The balance between innovation and ethics is delicate. What do you think? Should AI have a role in shaping our intellectual property? Let's discuss! Follow me for insights and DM me if you have exciting projects we can explore together. #generativeai #innovation #ethics #patents #technology

2

Accenture And NVIDIA Saddle Up To Bring #AI Computing Power To The Fore ______________ Mega foundation models, no matter how well trained they are on public data, won’t move the needle on the apps and agents that businesses compete on. To tap genAI’s potential as a differentiating toolkit, firms need to build and deploy cost-effective proprietary models trained on their data. Accenture is preparing to do that on NVIDIA’s stack of AI chips, models, and software (Llama-based models, too). This is an opening salvo in the battle for AI computing ecosystems to host the dozens or hundreds of proprietary models that a business will need. — Ted Schadler

7

1 Comment

What's the secret behind a great story? It may not be what you think. From the outline to timeline to promotion, a well-crated project plan will set you up for success every time. Learn the basics from the #StoryMaps team: https://ow.ly/Thsm50TScJ3

6

Envisioning Information Access Systems "The primary goal of this conceptual article is to shed light on what we still do not know about the potential impacts of #LLM-based information access system..." ~ @chirag_shah @emilymbender #AI #GenerativeAI

2

We have seen the LLM trend toward bigger and bigger models. Now some are taking the opposite approach to bring small targeted and cost effective models to the market. Maybe this will be an angle to help mitigate AI draining the power grid. #Innovation #MiniModel

9

1 Comment

Software people, 1) have you done an analysis of the data using to train your AI apps? 2) are you vetting the AI results (for truth-iness) before they are presented to the end user? https://lnkd.in/g4jsrS76 As I say in https://lnkd.in/g53xBXgw - “Garbage Data” leads to “Garbage AI Systems”

6

1 Comment

So as AI disrupts search we will see more of these. So in this example they have a AI search portal with a ton of LLMs behind. So I guess there must be some AI figuring out which AI to engage 🤫 Once the search also allows actions the traditional ad ridden search engines will be in deep trouble

10

2 Comments