ChatGPT Archives - gettectonic.com - Page 2
OpenAI Introduces Canvas
31Oct

OpenAI Introduces Canvas

Don’t get spooked – OpenAI introduces Canvas—a fresh interface for collaborative writing and coding with ChatGPT, designed to go beyond simple conversation. Canvas opens in a separate window, enabling you and ChatGPT to work on projects side by side, creating and refining ideas in real time. This early beta provides an entirely new way of collaborating with AI—combining conversation with the ability to edit and enhance content together. Built on GPT-4o, Canvas can be selected in the model picker during the beta phase. Starting today, we’re rolling it out to ChatGPT Plus and Team users globally, with Enterprise and Education users gaining access next week. Once out of beta, Canvas will be available to all ChatGPT Free users. Enhancing Collaboration with ChatGPT While ChatGPT’s chat interface works well for many tasks, projects requiring editing and iteration benefit from more. Canvas provides a workspace designed for such needs. Here, ChatGPT can better interpret your objectives, offering inline feedback and suggestions across entire projects—similar to a copy editor or code reviewer. You control every aspect in Canvas, from direct editing to leveraging shortcuts like adjusting text length, debugging code, or quickly refining writing. You can also revert to previous versions with Canvas’s back button. OpenAI Introduces Canvas Canvas opens automatically when ChatGPT detects an ideal scenario, or you can prompt it by typing “use Canvas” in your request to begin working collaboratively on an existing project. Writing Shortcuts Include: Coding in Canvas Canvas makes coding revisions more transparent, streamlining the iterative coding process. Track ChatGPT’s edits more clearly and take advantage of features that make debugging and revising code simpler. OpenAI Introduces Canvas to a world of new possibilities for truly developing and working with artificial intelligence. Coding Shortcuts Include: Training the Model to Collaborate GPT-4o has been optimized to act as a collaborative partner, understanding when to open a Canvas, make targeted edits, or fully rewrite content. Our team implemented core behaviors to support a seamless experience, including: These improvements are backed by over 20 internal automated evaluations and refined with synthetic data generation techniques, allowing us to enhance response quality and interaction without relying on human-generated data. Key Challenges as OpenAI Introduces Canvas A core challenge was determining when to trigger Canvas. We trained GPT-4o to recognize prompts like “Write a blog post about the history of coffee beans” while avoiding over-triggering for simple Q&A requests. For writing tasks, we reached an 83% accuracy in correct Canvas triggers, and a 94% accuracy in coding tasks compared to baseline models. Fine-tuning continues to ensure targeted edits are favored over full rewrites when needed. Finally, improving comment generation required iterative adjustments and human evaluations, with the integrated Canvas model now outperforming baseline GPT-4o in accuracy by 30% and quality by 16%. What’s Next Canvas is the first major update to ChatGPT’s visual interface since launch, with more enhancements planned to make AI more versatile and accessible. Canvas is also integrated with Salesforce. Like Related Posts Salesforce OEM AppExchange Expanding its reach beyond CRM, Salesforce.com has launched a new service called AppExchange OEM Edition, aimed at non-CRM service providers. Read more Salesforce Jigsaw Salesforce.com, a prominent figure in cloud computing, has finalized a deal to acquire Jigsaw, a wiki-style business contact database, for Read more Health Cloud Brings Healthcare Transformation Following swiftly after last week’s successful launch of Financial Services Cloud, Salesforce has announced the second installment in its series Read more Top Ten Reasons Why Tectonic Loves the Cloud The Cloud is Good for Everyone – Why Tectonic loves the cloud You don’t need to worry about tracking licenses. Read more

Read More
October 31, 2024in Data, Enterprise
More AI Tools to Use
29Oct

More AI Tools to Use

Additionally, Arc’s collaboration with Perplexity elevates browsing by transforming search experiences. Perplexity functions as a personal AI research assistant, fetching and summarizing information along with sources, visuals, and follow-up questions. Premium users even have access to advanced large language models like GPT-4 and Claude. Together, Arc and Perplexity revolutionize how users navigate the web. Like Related Posts Salesforce OEM AppExchange Expanding its reach beyond CRM, Salesforce.com has launched a new service called AppExchange OEM Edition, aimed at non-CRM service providers. Read more Salesforce Jigsaw Salesforce.com, a prominent figure in cloud computing, has finalized a deal to acquire Jigsaw, a wiki-style business contact database, for Read more Health Cloud Brings Healthcare Transformation Following swiftly after last week’s successful launch of Financial Services Cloud, Salesforce has announced the second installment in its series Read more Top Ten Reasons Why Tectonic Loves the Cloud The Cloud is Good for Everyone – Why Tectonic loves the cloud You don’t need to worry about tracking licenses. Read more

Read More
October 29, 2024in AI Tools, Data, Google
AI Agents and Tabular Data
28Oct

AI Agents and Tabular Data

Have you ever wondered how AI agents understand tabular data, such as that in CSV or Excel files? Or how a file loaded into a platform like ChatGPT can be instantly understood and processed? This insight explores the creation of a custom AI agent capable of achieving these tasks from scratch. AI Agents and Tabular Data. Context Jen, an AI Engineer at AI United, leads a team developing an AI agent within a 30-day timeline. This agent will generate tailored interactive charts based on uploaded data files, enabling users to better visualize and interpret the data. To achieve this, the team needed to ensure the AI agent could analyze the file’s data context and autonomously recommend the most appropriate chart types. The agent development was broken down into four main steps: Here’s a look at how the team developed the AI system to ingest CSV data and aggregate it into an actionable format. Setup The development began by configuring package installations and defining the language model to be used: pythonCopy code%pip install langchain_openai %pip install langchain_core %pip install langchain_community %pip install langchain_experimental from langchain_openai.chat_models import ChatOpenAI openai_key = os.environ.get(“OPENAI_API”) gpt4o = ChatOpenAI(temperature=0.0, model=”gpt-4o”, openai_api_key=openai_key) Step 1: Context Creation Before generating the code to process raw data, the team created context around the dataset to enhance the AI’s response accuracy. Metadata extraction included: For demonstration, a wine reviews dataset was used, and metadata was extracted as follows: pythonCopy codeimport pandas as pd def extract_metadata(df): metadata = { ‘Number of Columns’: df.shape[1], ‘Schema’: df.columns.tolist(), ‘Data Types’: str(df.dtypes), ‘Sample’: df.head(1).to_dict(orient=”records”) } return metadata df = pd.read_csv(“wine_reviews.csv”) metadata = extract_metadata(df) Step 2: Prompt Augmentation To help the AI model interpret the dataset, prompts were augmented with extracted metadata using a template: pythonCopy codeprompt_template = ”’ Assistant is an AI model that suggests charts to visualize data based on the following metadata. SCHEMA: ——– {schema} DATA TYPES: ——– {data_types} SAMPLE: ——– {sample} ”’.format(schema=metadata[“Schema”], data_types=metadata[“Data Types”], sample=metadata[“Sample”]) gpt40.invoke(prompt_template) Step 3: Simple Agent Code Generation & Execution With the prompt augmented, the model was able to suggest suitable charts. For data transformation, an agentic workflow with a Python REPL tool was used, where the AI generated code for aggregating data and then executed it to provide the necessary structure for plotting. A REPL instance was created to pass data into Python functions, enabling the AI to perform aggregation. pythonCopy codefrom langchain_experimental.utilities import PythonREPL repl = PythonREPL() repl.globals[‘df’] = df from langchain_core.tools import tool @tool def python_repl(code: str): try: result = repl.run(code) except BaseException as e: return f”Failed to execute. Error: {repr(e)}” result_str = f”Successfully executed:n“`pythonn{code}n“`nStdout: {result}” return result_str tools = [python_repl] Step 4: Final Data Aggregation and Charting Finally, the AI suggested the Bar Chart type for plotting the top 10 wineries by average points, and the REPL instance executed the code to transform the data for the chart: pythonCopy code# Code to aggregate and convert data into dictionary format top_wineries = df.groupby(‘winery’)[‘points’].mean().sort_values(ascending=False).head(10) top_wineries_dict = top_wineries.to_dict() print(top_wineries_dict) The aggregated data was output as: jsonCopy code{‘Macauley’: 96.0, ‘Heitz’: 95.5, ‘Bodega Carmen Rodríguez’: 95.5, ‘Maurodos’: 95.0, ‘Blue Farm’: 95.0, ‘Numanthia’: 95.0, ‘Château Lagrézette’: 95.0, ‘Patricia Green Cellars’: 95.0, ‘Ponzi’: 95.0, ‘Muga’: 95.0} With this approach, the AI agent was capable of not only understanding the data in the uploaded file but also generating interactive visualizations, making complex datasets more accessible and insightful. Like Related Posts Salesforce OEM AppExchange Expanding its reach beyond CRM, Salesforce.com has launched a new service called AppExchange OEM Edition, aimed at non-CRM service providers. Read more Salesforce Jigsaw Salesforce.com, a prominent figure in cloud computing, has finalized a deal to acquire Jigsaw, a wiki-style business contact database, for Read more Health Cloud Brings Healthcare Transformation Following swiftly after last week’s successful launch of Financial Services Cloud, Salesforce has announced the second installment in its series Read more Top Ten Reasons Why Tectonic Loves the Cloud The Cloud is Good for Everyone – Why Tectonic loves the cloud You don’t need to worry about tracking licenses. Read more

Read More
October 28, 2024in Agentforce Platform, Artificial Intelligence, Data, Generative AI, Technology
AI FOMO
26Oct

AI FOMO

Enterprise interest in artificial intelligence has surged in the past two years, with boardroom discussions centered on how to capitalize on AI advancements before competitors do. Generative AI has been a particular focus for executives since the launch of ChatGPT in November 2022, followed by other major product releases like Amazon’s Bedrock, Google’s Gemini, Meta’s Llama, and a host of SaaS tools incorporating the technology. However, the initial rush driven by fear of missing out (FOMO) is beginning to fade. Business and tech leaders are now shifting their attention from experimentation to more practical concerns: How can AI generate revenue? This question will grow in importance as pilot AI projects move into production, raising expectations for financial returns. Using AI to Increase Revenue AI’s potential to drive revenue will be a critical factor in determining how quickly organizations adopt the technology and how willing they are to invest further. Here are 10 ways businesses can harness AI to boost revenue: 1. Boost Sales AI-powered virtual assistants and chatbots can help increase sales. For example, Ikea’s generative AI tool assists customers in designing their living spaces while shopping for furniture. Similarly, jewelry insurance company BriteCo launched a GenAI chatbot that reduced chat abandonment rates, leading to more successful customer interactions and potentially higher sales. A TechTarget survey revealed that AI-powered customer-facing tools like chatbots are among the top investments for IT leaders. 2. Reduce Customer Churn AI helps businesses retain clients, reducing revenue loss and improving customer lifetime value. By analyzing historical data, AI can profile customer attributes and identify accounts at risk of leaving. AI can then assist in personalizing customer experiences, decreasing churn and fostering loyalty. 3. Enhance Recommendation Engines AI algorithms can analyze customer data to offer personalized product recommendations. This drives cross-selling and upselling opportunities, boosting revenue. For instance, Meta’s AI-powered recommendation engine has increased user engagement across its platforms, attracting more advertisers. 4. Accelerate Marketing Strategies While marketing doesn’t directly generate revenue, it fuels the sales pipeline. Generative AI can quickly produce personalized content, such as newsletters and ads, tailored to customer interests. Gartner predicts that by 2025, 30% of outbound marketing messages will be AI-generated, up from less than 2% in 2022. 5. Detect Fraud AI is instrumental in detecting fraudulent activities, helping businesses preserve revenue. Financial firms like Capital One use machine learning to detect anomalies and prevent credit card fraud, while e-commerce companies leverage AI to flag fraudulent orders. 6. Reinvent Business Processes AI can transform entire business processes, unlocking new revenue streams. For example, Accenture’s 2024 report highlighted an insurance company that expects a 10% revenue boost after retooling its underwriting workflow with AI. In healthcare, AI could streamline revenue cycle management, speeding up reimbursement processes. 7. Develop New Products and Services AI accelerates product development, particularly in industries like pharmaceuticals, where it assists in drug discovery. AI tools also speed up the delivery of digital products, as seen with companies like Ally Financial and ServiceNow, which have reduced software development times by 20% or more. 8. Provide Predictive Maintenance AI-driven predictive maintenance helps prevent costly equipment downtime in industries like manufacturing and fleet management. By identifying equipment on the brink of failure, AI allows companies to schedule repairs and avoid revenue loss from operational disruptions. 9. Improve Forecasting AI’s predictive capabilities enhance planning and forecasting. By analyzing historical and real-time data, AI can predict product demand and customer behavior, enabling businesses to optimize inventory levels and ensure product availability for ready-to-buy customers. 10. Optimize Pricing AI can dynamically adjust prices based on factors like demand shifts and competitor pricing. Reinforcement learning algorithms allow businesses to optimize pricing in real time, ensuring they maximize revenue even as market conditions change. Keeping ROI in Focus While AI offers numerous ways to generate new revenue streams, it also introduces costs in development, infrastructure, and operations—some of which may not be immediately apparent. For instance, research from McKinsey & Company shows that GenAI models account for only 15% of a project’s total cost, with additional expenses related to change management and data preparation often overlooked. To make the most of AI, organizations should prioritize use cases with a clear return on investment (ROI) and postpone those that don’t justify the expense. A focus on ROI ensures that AI deployments align with business goals and contribute to sustainable revenue growth. Like Related Posts Salesforce OEM AppExchange Expanding its reach beyond CRM, Salesforce.com has launched a new service called AppExchange OEM Edition, aimed at non-CRM service providers. Read more The Salesforce Story In Marc Benioff’s own words How did salesforce.com grow from a start up in a rented apartment into the world’s Read more Salesforce Jigsaw Salesforce.com, a prominent figure in cloud computing, has finalized a deal to acquire Jigsaw, a wiki-style business contact database, for Read more Health Cloud Brings Healthcare Transformation Following swiftly after last week’s successful launch of Financial Services Cloud, Salesforce has announced the second installment in its series Read more

Read More
October 26, 2024in Generative AI
Scope of Generative AI
24Oct

Scope of Generative AI

Generative AI has far more to offer your site than simply mimicking a conversational ChatGPT-like experience or providing features like generating cover letters on resume sites. Let’s explore how you can integrate Generative AI with your product in diverse and innovative ways! There are three key perspectives to consider when integrating Generative AI with your features: system scope, spatial relationship, and functional relationship. Each perspective offers a different lens for exploring integration pathways and can spark valuable conversations about melding AI with your product ecosystem. These categories aren’t mutually exclusive; instead, they overlap and provide flexible ways of envisioning AI’s role. 1. System Scope — The Reach of Generative AI in Your System System scope refers to the breadth of integration within your system. By viewing integration from this angle, you can assess the role AI plays in managing your platform’s overall functionality. While these categories may overlap, they are useful in facilitating strategic conversations. 2. Spatial Relationships — Where AI Interacts with Features Spatial relationships describe where AI features sit in relation to your platform’s functionality: 3. Functional Relationships — How AI Interacts with Features Functional relationships determine how AI and platform features work together. This includes how users engage with AI and how AI content updates based on feature interactions: Scope of Generative AI By considering these different perspectives—system scope, spatial, and functional—you can drive more meaningful conversations about how Generative AI can best enhance your product’s capabilities. Each approach offers unique value, and careful thought can help teams choose the integration path that aligns with their needs and goals. Scope of Generative AI conversations with Tectonic can assist in planning the best ROI approach to AI. Contact us today. Like Related Posts Salesforce OEM AppExchange Expanding its reach beyond CRM, Salesforce.com has launched a new service called AppExchange OEM Edition, aimed at non-CRM service providers. Read more Salesforce Jigsaw Salesforce.com, a prominent figure in cloud computing, has finalized a deal to acquire Jigsaw, a wiki-style business contact database, for Read more Health Cloud Brings Healthcare Transformation Following swiftly after last week’s successful launch of Financial Services Cloud, Salesforce has announced the second installment in its series Read more Top Ten Reasons Why Tectonic Loves the Cloud The Cloud is Good for Everyone – Why Tectonic loves the cloud You don’t need to worry about tracking licenses. Read more

Read More
October 24, 2024in Data, Generative AI, Google
Scaling Generative AI
23Oct

Scaling Generative AI

Many organizations follow a hybrid approach to AI infrastructure, combining public clouds, colocation facilities, and on-prem solutions. Specialized GPU-as-a-service vendors, for instance, are becoming popular for handling high-demand AI computations, helping businesses manage costs without compromising performance. Business process outsourcing company TaskUs, for example, focuses on optimizing compute and data flows as it scales its gen AI deployments, while Cognizant advises that companies distinguish between training and inference needs, each with different latency requirements.

Read More
October 23, 2024in Data, Enterprise, Generative AI, Google, Salesforce
AI Agents and Digital Transformation
23Oct

Ready for AI Agents

Brands that can effectively integrate agentic AI into their operations stand to gain a significant competitive edge. But as with any innovation, success will depend on balancing the promise of automation with the complexities of trust, privacy, and user experience.

Read More
October 23, 2024in Data, Enterprise, Generative AI, Google, Salesforce
AI in Programming
23Oct

AI in Programming

Since the launch of ChatGPT in 2022, developers have been split into two camps: those who ban AI in coding and those who embrace it. Many seasoned programmers not only avoid AI-generated code but also prohibit their teams from using it. Their reasoning is simple: “AI-generated code is unreliable.” Even if one doesn’t agree with this anti-AI stance, they’ve likely faced challenges, hurdles, or frustrations when using AI for programming. The key is finding the right strategies to use AI to your advantage. Many are still using outdated AI strategies from two years ago, likened to cutting down a tree with kitchen knives. Two Major Issues with AI for Developers The Wrong Way to Use AI… …can be broken down into two parts: When ChatGPT first launched, the typical way to work with AI was to visit the website and chat with GPT-3.5 in a browser. The process was straightforward: copy code from the IDE, paste it into ChatGPT with a basic prompt like “add comments,” get the revised code, check for errors, and paste it back into the IDE. Many developers, especially beginners and students, are still using this same method. However, the AI landscape has changed significantly over the last two years, and many have not adjusted their approach to fully leverage AI’s potential. Another common pitfall is how developers use AI. They ask the LLM to generate code, test it, and go back and forth to fix any issues. Often, they fall into an endless loop of AI hallucinations when trying to get the LLM to understand what’s wrong. This can be frustrating and unproductive. Four Tools to Boost Programming Productivity with AI 1. Cursor: AI-First IDE Cursor is an AI-first IDE built on VScode but enhanced with AI features. It allows developers to integrate a chatbot API and use AI as an assistant. Some of Cursor’s standout features include: Cursor integrates seamlessly with VScode, making it easy for existing users to transition. It supports various models, including GPT-4, Claude 3.5 Sonnet, and its built-in free cursor-small model. The combination of Cursor and Sonnet 3.5 has been particularly praised for producing reliable coding results. This tool is a significant improvement over copy-pasting code between ChatGPT and an IDE. 2. Micro Agent: Code + Test Case Micro Agent takes a different approach to AI-generated code by focusing on test cases. Instead of generating large chunks of code, it begins by creating test cases based on the prompt, then writes code that passes those tests. This method results in more grounded and reliable output, especially for functions that are tricky but not overly complex. 3. SWE-agent: AI for GitHub Issues Developed by Princeton Language and Intelligence, SWE-agent specializes in resolving real-world GitHub repository issues and submitting pull requests. It’s a powerful tool for managing large repositories, as it reviews codebases, identifies issues, and makes necessary changes. SWE-agent is open-source and has gained considerable popularity on GitHub. 4. AI Commits: git commit -m AI Commits generates meaningful commit messages based on your git diff. This simple tool eliminates the need for vague or repetitive commit messages like “minor changes.” It’s easy to install and uses GPT-3.5 for efficient, AI-generated commit messages. The Path Forward To stay productive and achieve goals in the rapidly evolving AI landscape, developers need the right tools. The limitations of AI, such as hallucinations, can’t be eliminated, but using the appropriate tools can help mitigate them. Simple, manual interactions like generating code or comments through ChatGPT can be frustrating. By adopting the right strategies and tools, developers can avoid these pitfalls and confidently enhance their coding practices. AI is evolving fast, and keeping up with its changes is crucial. The right tools can make all the difference in your programming workflow. Like Related Posts Salesforce OEM AppExchange Expanding its reach beyond CRM, Salesforce.com has launched a new service called AppExchange OEM Edition, aimed at non-CRM service providers. Read more The Salesforce Story In Marc Benioff’s own words How did salesforce.com grow from a start up in a rented apartment into the world’s Read more Salesforce Jigsaw Salesforce.com, a prominent figure in cloud computing, has finalized a deal to acquire Jigsaw, a wiki-style business contact database, for Read more Health Cloud Brings Healthcare Transformation Following swiftly after last week’s successful launch of Financial Services Cloud, Salesforce has announced the second installment in its series Read more

Read More
October 23, 2024in Generative AI, Salesforce
AI Risk Management
22Oct

AI Risk Management

Organizations must acknowledge the risks associated with implementing AI systems to use the technology ethically and minimize liability. Throughout history, companies have had to manage the risks associated with adopting new technologies, and AI is no exception. Some AI risks are similar to those encountered when deploying any new technology or tool, such as poor strategic alignment with business goals, a lack of necessary skills to support initiatives, and failure to secure buy-in across the organization. For these challenges, executives should rely on best practices that have guided the successful adoption of other technologies. In the case of AI, this includes: However, AI introduces unique risks that must be addressed head-on. Here are 15 areas of concern that can arise as organizations implement and use AI technologies in the enterprise: Managing AI Risks While AI risks cannot be eliminated, they can be managed. Organizations must first recognize and understand these risks and then implement policies to minimize their negative impact. These policies should ensure the use of high-quality data, require testing and validation to eliminate biases, and mandate ongoing monitoring to identify and address unexpected consequences. Furthermore, ethical considerations should be embedded in AI systems, with frameworks in place to ensure AI produces transparent, fair, and unbiased results. Human oversight is essential to confirm these systems meet established standards. For successful risk management, the involvement of the board and the C-suite is crucial. As noted, “This is not just an IT problem, so all executives need to get involved in this.” Like Related Posts Salesforce OEM AppExchange Expanding its reach beyond CRM, Salesforce.com has launched a new service called AppExchange OEM Edition, aimed at non-CRM service providers. Read more Salesforce Jigsaw Salesforce.com, a prominent figure in cloud computing, has finalized a deal to acquire Jigsaw, a wiki-style business contact database, for Read more Health Cloud Brings Healthcare Transformation Following swiftly after last week’s successful launch of Financial Services Cloud, Salesforce has announced the second installment in its series Read more Alphabet Soup of Cloud Terminology As with any technology, the cloud brings its own alphabet soup of terms. This insight will hopefully help you navigate Read more

Read More
October 22, 2024in AI Tools, Data, Enterprise, Generative AI, Laws, Technology
What is Explainable AI
22Oct

What is Explainable AI

Building a trusted AI system starts with ensuring transparency in how decisions are made. Explainable AI is vital not only for addressing trust issues within organizations but also for navigating regulatory challenges. According to research from Forrester, many business leaders express concerns over AI, particularly generative AI, which surged in popularity following the 2022 release of ChatGPT by OpenAI. “AI faces a trust issue,” explained Forrester analyst Brandon Purcell, underscoring the need for explainability to foster accountability. He highlighted that explainability helps stakeholders understand how AI systems generate their outputs. “Explainability builds trust,” Purcell stated at the Forrester Technology and Innovation Summit in Austin, Texas. “When employees trust AI systems, they’re more inclined to use them.” Implementing explainable AI does more than encourage usage within an organization—it also helps mitigate regulatory risks, according to Purcell. Explainability is crucial for compliance, especially under regulations like the EU AI Act. Forrester analyst Alla Valente emphasized the importance of integrating accountability, trust, and security into AI efforts. “Don’t wait for regulators to set standards—ensure you’re already meeting them,” she advised at the summit. Purcell noted that explainable AI varies depending on whether the AI model is predictive, generative, or agentic. Building an Explainable AI System AI explainability encompasses several key elements, including reproducibility, observability, transparency, interpretability, and traceability. For predictive models, transparency and interpretability are paramount. Transparency involves using “glass-box modeling,” where users can see how the model analyzed the data and arrived at its predictions. This approach is likely to be a regulatory requirement, especially for high-risk applications. Interpretability is another important technique, useful for lower-risk cases such as fraud detection or explaining loan decisions. Techniques like partial dependence plots show how specific inputs affect predictive model outcomes. “With predictive AI, explainability focuses on the model itself,” Purcell noted. “It’s one area where you can open the hood and examine how it works.” In contrast, generative AI models are often more opaque, making explainability harder. Businesses can address this by documenting the entire system, a process known as traceability. For those using models from vendors like Google or OpenAI, tools like transparency indexes and model cards—which detail the model’s use case, limitations, and performance—are valuable resources. Lastly, for agentic AI systems, which autonomously pursue goals, reproducibility is key. Businesses must ensure that the model’s outputs can be consistently replicated with similar inputs before deployment. These systems, like self-driving cars, will require extensive testing in controlled environments before being trusted in the real world. “Agentic systems will need to rack up millions of virtual miles before we let them loose,” Purcell concluded. Like Related Posts Salesforce OEM AppExchange Expanding its reach beyond CRM, Salesforce.com has launched a new service called AppExchange OEM Edition, aimed at non-CRM service providers. Read more Salesforce Jigsaw Salesforce.com, a prominent figure in cloud computing, has finalized a deal to acquire Jigsaw, a wiki-style business contact database, for Read more Health Cloud Brings Healthcare Transformation Following swiftly after last week’s successful launch of Financial Services Cloud, Salesforce has announced the second installment in its series Read more Top Ten Reasons Why Tectonic Loves the Cloud The Cloud is Good for Everyone – Why Tectonic loves the cloud You don’t need to worry about tracking licenses. Read more

Read More
October 22, 2024in Data, Generative AI, Google, Technology
AI in Networking
21Oct

AI in Networking

AI Tools in Networking: Tailoring Capabilities to Unique Needs AI tools are becoming increasingly common across various industries, offering a wide range of functionalities. However, network engineers may not require every capability these tools provide. Each network has distinct requirements that align with specific business objectives, necessitating that network engineers and developers select AI toolsets tailored to their networks’ needs. While network teams often desire similar AI capabilities, they also encounter common challenges in integrating these tools into their systems. The Rise of AI in Networking Though AI is not a new concept—having existed for decades in the form of automated and expert systems—it is gaining unprecedented attention. According to Jim Frey, principal analyst for networking at TechTarget’s Enterprise Strategy Group, many organizations have not fully grasped AI’s potential in production environments over the past three years. “AI has been around for a long time, but the interesting thing is, only a minority—not even half—have really said they’re using it effectively in production for the last three years,” Frey noted. Generative AI (GenAI) has significantly contributed to this renewed interest in AI. Shamus McGillicuddy, vice president of research at Enterprise Management Associates, categorizes AI tools into two main types: GenAI and AIOps (AI for IT operations). “Generative AI, like ChatGPT, has recently surged in popularity, becoming a focal point of discussion among IT professionals,” McGillicuddy explained. “AIOps, on the other hand, encompasses machine learning, anomaly detection, and analytics.” The increasing complexity of networks is another factor driving the adoption of AI in networking. Frey highlighted that the demands of modern network environments are beyond human capability to manage manually, making AI engines a vital solution. Essential AI Tool Capabilities for Networks While individual network needs vary, many network engineers seek similar functionalities when integrating AI. Commonly desired capabilities include: According to McGillicuddy’s research, network optimization and automated troubleshooting are among the most popular use cases for AI. However, many professionals prefer to retain manual oversight in the fixing process. “Automated troubleshooting can identify and analyze issues, but typically, people want to approve the proposed fixes,” McGillicuddy stated. Many of these capabilities are critical for enhancing security and mitigating threats. Frey emphasized that networking professionals increasingly view AI as a tool to improve organizational security. DeCarlo echoed this sentiment, noting that network managers share similar objectives with security professionals regarding proactive problem recognition. Frey also mentioned alternative use cases for AI, such as documentation and change recommendations, which, while less popular, can offer significant value to network teams. Ultimately, the relevance of any AI capability hinges on its fit within the network environment and team needs. “I don’t think you can prioritize one capability over another,” DeCarlo remarked. “It depends on the tools being used and their effectiveness.” Generative AI: A New Frontier Despite its recent emergence, GenAI has quickly become an asset in the networking field. McGillicuddy noted that in the past year and a half, network professionals have adopted GenAI tools, with ChatGPT being one of the most recognized examples. “One user reported that leveraging ChatGPT could reduce a task that typically takes four hours down to just 10 minutes,” McGillicuddy said. However, he cautioned that users must understand the limitations of GenAI, as mistakes can occur. “There’s a risk of errors or ‘hallucinations’ with these tools, and having blind faith in their outputs can lead to significant network issues,” he warned. In addition to ChatGPT, vendors are developing GenAI interfaces for their products, including virtual assistants. According to McGillicuddy’s findings, common use cases for vendor GenAI products include: DeCarlo added that GenAI tools offer valuable training capabilities due to their rapid processing speeds and in-depth analysis, which can expedite knowledge acquisition within the network. Frey highlighted that GenAI’s rise is attributed to its ability to outperform older systems lacking sophistication. Nevertheless, the complexity of GenAI infrastructures has led to a demand for AIOps tools to manage these systems effectively. “We won’t be able to manage GenAI infrastructures without the support of AI tools, as human capabilities cannot keep pace with rapid changes,” Frey asserted. Challenges in Implementing AI Tools While AI tools present significant benefits for networks, network engineers and managers must navigate several challenges before integration. Data Privacy, Collection, and Quality Data usage remains a critical concern for organizations considering AIOps and GenAI tools. Frey noted that the diverse nature of network data—combining operational information with personally identifiable information—heightens data privacy concerns. For GenAI, McGillicuddy pointed out the importance of validating AI outputs and ensuring high-quality data is utilized for training. “If you feed poor data to a generative AI tool, it will struggle to accurately understand your network,” he explained. Complexity of AI Tools Frey and McGillicuddy agreed that the complexity of both AI and network systems could hinder effective deployment. Frey mentioned that AI systems, especially GenAI, require careful tuning and strong recommendations to minimize inaccuracies. McGillicuddy added that intricate network infrastructures, particularly those involving multiple vendors, could limit the effectiveness of AIOps components, which are often specialized for specific systems. User Uptake and Skills Gaps User adoption of AI tools poses a significant challenge. Proper training is essential to realize the full benefits of AI in networking. Some network professionals may be resistant to using AI, while others may lack the knowledge to integrate these tools effectively. McGillicuddy noted that AIOps tools are often less intuitive than GenAI, necessitating a certain level of expertise for users to extract value. “Understanding how tools function and identifying potential gaps can be challenging,” DeCarlo added. The learning curve can be steep, particularly for teams accustomed to longstanding tools. Integration Issues Integration challenges can further complicate user adoption. McGillicuddy highlighted two dimensions of this issue: tools and processes. On the tools side, concerns arise about harmonizing GenAI with existing systems. “On the process side, it’s crucial to ensure that teams utilize these tools effectively,” he said. DeCarlo cautioned that organizations might need to create in-house supplemental tools to bridge integration gaps, complicating the synchronization of vendor AI

Read More
October 21, 2024in AI Tools, Data, Enterprise, Generative AI
Fully Formatted Facts
20Oct

Fully Formatted Facts

A recent discovery by programmer and inventor Michael Calvin Wood is addressing a persistent challenge in AI: hallucinations. These false or misleading outputs, long considered an inherent flaw in large language models (LLMs), have posed a significant issue for developers. However, Wood’s breakthrough is challenging this assumption, offering a solution that could transform how AI-powered applications are built and used. The Importance of Wood’s Discovery for Developers Wood’s findings have substantial implications for developers working with AI. By eliminating hallucinations, developers can ensure that AI-generated content is accurate and reliable, particularly in applications where precision is critical. Understanding the Root Cause of Hallucinations Contrary to popular belief, hallucinations are not primarily caused by insufficient training data or biased algorithms. Wood’s research reveals that the issue stems from how LLMs process and generate information based on “noun-phrase routes.” LLMs organize information around noun phrases, and when they encounter semantically similar phrases, they may conflate or misinterpret them, leading to incorrect outputs. How LLMs Organize Information For example: The Noun-Phrase Dominance Model Wood’s research led to the development of the Noun-Phrase Dominance Model, which posits that neural networks in LLMs self-organize around noun phrases. This model is key to understanding and eliminating hallucinations by addressing how AI processes noun-phrase conflicts. Fully-Formatted Facts (FFF): A Solution Wood’s solution involves transforming input data into Fully-Formatted Facts (FFF)—statements that are literally true, devoid of noun-phrase conflicts, and structured as simple, complete sentences. Presenting information in this format has led to significant improvements in AI accuracy, particularly in question-answering tasks. How FFF Processing Works While Wood has not provided a step-by-step guide for FFF processing, he hints that the process began with named-entity recognition using the Python SpaCy library and evolved into using an LLM to reduce ambiguity while retaining the original writing style. His company’s REST API offers a wrapper around GPT-4o and GPT-4o-mini models, transforming input text to remove ambiguity before processing it. Current Methods vs. Wood’s Approach Current approaches, like Retrieval Augmented Generation (RAG), attempt to reduce hallucinations by adding more context. However, these methods often introduce additional noun-phrase conflicts. For instance, even with RAG, ChatGPT-3.5 Turbo experienced a 23% hallucination rate when answering questions about Wikipedia articles. In contrast, Wood’s method focuses on eliminating noun-phrase conflicts entirely. Results: RAG FF (Retrieval Augmented Generation with Formatted Facts) Wood’s method has shown remarkable results, eliminating hallucinations in GPT-4 and GPT-3.5 Turbo during question-answering tasks using third-party datasets. Real-World Example: Translation Error Elimination Consider a simple translation example: This transformation eliminates hallucinations by removing the potential noun-phrase conflict. Implications for the Future of AI The Noun-Phrase Dominance Model and the use of Fully-Formatted Facts have far-reaching implications: Roadmap for Future Development Wood and his team plan to expand their approach by: Conclusion: A New Era of Reliable AI Wood’s discovery represents a significant leap forward in the pursuit of reliable AI. By aligning input data with how LLMs process information, he has unlocked the potential for accurate, trustworthy AI systems. As this technology continues to evolve, it could have profound implications for industries ranging from healthcare to legal services, where AI could become a consistent and reliable tool. While there is still work to be done in expanding this method across all AI tasks, the foundation has been laid for a revolution in AI accuracy. Future developments will likely focus on refining and expanding these capabilities, enabling AI to serve as a trusted resource across a range of applications. Experience RAGFix For those looking to explore this technology, RAGFix offers an implementation of these groundbreaking concepts. Visit their official website to access demos, explore REST API integration options, and stay updated on the latest advancements in hallucination-free AI: Visit RAGFix.ai Like Related Posts Salesforce OEM AppExchange Expanding its reach beyond CRM, Salesforce.com has launched a new service called AppExchange OEM Edition, aimed at non-CRM service providers. Read more The Salesforce Story In Marc Benioff’s own words How did salesforce.com grow from a start up in a rented apartment into the world’s Read more Salesforce Jigsaw Salesforce.com, a prominent figure in cloud computing, has finalized a deal to acquire Jigsaw, a wiki-style business contact database, for Read more Health Cloud Brings Healthcare Transformation Following swiftly after last week’s successful launch of Financial Services Cloud, Salesforce has announced the second installment in its series Read more

Read More
October 20, 2024in Salesforce
AI Prompts to Accelerate Academic Reading
19Oct

AI Prompts to Accelerate Academic Reading

10 AI Prompts to Accelerate Academic Reading with ChatGPT and Claude AI In the era of information overload, keeping pace with academic research can feel daunting. Tools like ChatGPT and Claude AI can streamline your reading and help you extract valuable insights from research papers quickly and efficiently. These AI assistants, when used ethically and responsibly, support your critical analysis by summarizing complex studies, highlighting key findings, and breaking down methodologies. While these prompts enhance efficiency, they should complement—never replace—your own critical thinking and thorough reading. AI Prompts for Academic Reading 1. Elevator Pitch Summary Prompt: “Summarize this paper in 3–5 sentences as if explaining it to a colleague during an elevator ride.”This prompt distills the essence of a paper, helping you quickly grasp the core idea and decide its relevance. 2. Key Findings Extraction Prompt: “List the top 5 key findings or conclusions from this paper, with a brief explanation of each.”Cut through jargon to access the research’s core contributions in seconds. 3. Methodology Breakdown Prompt: “Explain the study’s methodology in simple terms. What are its strengths and potential limitations?”Understand the foundation of the research and critically evaluate its validity. 4. Literature Review Assistant Prompt: “Identify the key papers cited in the literature review and summarize each in one sentence, explaining its connection to the study.”A game-changer for understanding the context and building your own literature review. 5. Jargon Buster Prompt: “List specialized terms or acronyms in this paper with definitions in plain language.”Create a personalized glossary to simplify dense academic language. 6. Visual Aid Interpreter Prompt: “Explain the key takeaways from Figure X (or Table Y) and its significance to the study.”Unlock insights from charts and tables, ensuring no critical information is missed. 7. Implications Explorer Prompt: “What are the potential real-world implications or applications of this research? Suggest 3–5 possible impacts.”Connect theory to practice by exploring broader outcomes and significance. 8. Cross-Disciplinary Connections Prompt: “How might this paper’s findings or methods apply to [insert your field]? Suggest potential connections or applications.”Encourage interdisciplinary thinking by finding links between research areas. 9. Future Research Generator Prompt: “Based on the limitations and unanswered questions, suggest 3–5 potential directions for future research.”Spark new ideas and identify gaps for exploration in your field. 10. The Devil’s Advocate Prompt: “Play devil’s advocate: What criticisms or counterarguments could be made against the paper’s main claims? How might the authors respond?”Refine your critical thinking and prepare for discussions or reviews. Additional Resources Generative AI Prompts with Retrieval Augmented GenerationAI Agents and Tabular DataAI Evolves With Agentforce and Atlas Conclusion Incorporating these prompts into your routine can help you process information faster, understand complex concepts, and uncover new insights. Remember, AI is here to assist—not replace—your research skills. Stay critical, adapt prompts to your needs, and maximize your academic productivity. Like Related Posts Salesforce OEM AppExchange Expanding its reach beyond CRM, Salesforce.com has launched a new service called AppExchange OEM Edition, aimed at non-CRM service providers. Read more Salesforce Jigsaw Salesforce.com, a prominent figure in cloud computing, has finalized a deal to acquire Jigsaw, a wiki-style business contact database, for Read more Health Cloud Brings Healthcare Transformation Following swiftly after last week’s successful launch of Financial Services Cloud, Salesforce has announced the second installment in its series Read more Alphabet Soup of Cloud Terminology As with any technology, the cloud brings its own alphabet soup of terms. This insight will hopefully help you navigate Read more

Read More
October 19, 2024in AI Tools, Data, Generative AI
Recent advancements in AI
18Oct

Recent advancements in AI

Recent advancements in AI have been propelled by large language models (LLMs) containing billions to trillions of parameters. Parameters—variables used to train and fine-tune machine learning models—have played a key role in the development of generative AI. As the number of parameters grows, models like ChatGPT can generate human-like content that was unimaginable just a few years ago. Parameters are sometimes referred to as “features” or “feature counts.” While it’s tempting to equate the power of AI models with their parameter count, similar to how we think of horsepower in cars, more parameters aren’t always better. An increase in parameters can lead to additional computational overhead and even problems like overfitting. There are various ways to increase the number of parameters in AI models, but not all approaches yield the same improvements. For example, Google’s Switch Transformers scaled to trillions of parameters, but some of their smaller models outperformed them in certain use cases. Thus, other metrics should be considered when evaluating AI models. The exact relationship between parameter count and intelligence is still debated. John Blankenbaker, principal data scientist at SSA & Company, notes that larger models tend to replicate their training data more accurately, but the belief that more parameters inherently lead to greater intelligence is often wishful thinking. He points out that while these models may sound knowledgeable, they don’t actually possess true understanding. One challenge is the misunderstanding of what a parameter is. It’s not a word, feature, or unit of data but rather a component within the model‘s computation. Each parameter adjusts how the model processes inputs, much like turning a knob in a complex machine. In contrast to parameters in simpler models like linear regression, which have a clear interpretation, parameters in LLMs are opaque and offer no insight on their own. Christine Livingston, managing director at Protiviti, explains that parameters act as weights that allow flexibility in the model. However, more parameters can lead to overfitting, where the model performs well on training data but struggles with new information. Adnan Masood, chief AI architect at UST, highlights that parameters influence precision, accuracy, and data management needs. However, due to the size of LLMs, it’s impractical to focus on individual parameters. Instead, developers assess models based on their intended purpose, performance metrics, and ethical considerations. Understanding the data sources and pre-processing steps becomes critical in evaluating the model’s transparency. It’s important to differentiate between parameters, tokens, and words. A parameter is not a word; rather, it’s a value learned during training. Tokens are fragments of words, and LLMs are trained on these tokens, which are transformed into embeddings used by the model. The number of parameters influences a model’s complexity and capacity to learn. More parameters often lead to better performance, but they also increase computational demands. Larger models can be harder to train and operate, leading to slower response times and higher costs. In some cases, smaller models are preferred for domain-specific tasks because they generalize better and are easier to fine-tune. Transformer-based models like GPT-4 dwarf previous generations in parameter count. However, for edge-based applications where resources are limited, smaller models are preferred as they are more adaptable and efficient. Fine-tuning large models for specific domains remains a challenge, often requiring extensive oversight to avoid problems like overfitting. There is also growing recognition that parameter count alone is not the best way to measure a model’s performance. Alternatives like Stanford’s HELM and benchmarks such as GLUE and SuperGLUE assess models across multiple factors, including fairness, efficiency, and bias. Three trends are shaping how we think about parameters. First, AI developers are improving model performance without necessarily increasing parameters. A study of 231 models between 2012 and 2023 found that the computational power required for LLMs has halved every eight months, outpacing Moore’s Law. Second, new neural network approaches like Kolmogorov-Arnold Networks (KANs) show promise, achieving comparable results to traditional models with far fewer parameters. Lastly, agentic AI frameworks like Salesforce’s Agentforce offer a new architecture where domain-specific AI agents can outperform larger general-purpose models. As AI continues to evolve, it’s clear that while parameter count is an important consideration, it’s just one of many factors in evaluating a model’s overall capabilities. To stay on the cutting edge of artificial intelligence, contact Tectonic today. Like Related Posts Salesforce OEM AppExchange Expanding its reach beyond CRM, Salesforce.com has launched a new service called AppExchange OEM Edition, aimed at non-CRM service providers. Read more The Salesforce Story In Marc Benioff’s own words How did salesforce.com grow from a start up in a rented apartment into the world’s Read more Salesforce Jigsaw Salesforce.com, a prominent figure in cloud computing, has finalized a deal to acquire Jigsaw, a wiki-style business contact database, for Read more Health Cloud Brings Healthcare Transformation Following swiftly after last week’s successful launch of Financial Services Cloud, Salesforce has announced the second installment in its series Read more

Read More
October 18, 2024in Data, Generative AI, Google, Salesforce
NYT Issues Cease-and-Desist Letter to Perplexity AI
16Oct

NYT Issues Cease-and-Desist Letter to Perplexity AI

NYT Issues Cease-and-Desist Letter to Perplexity AI Over Alleged Unauthorized Content Use The New York Times (NYT) has issued a cease-and-desist letter to Perplexity AI, accusing the AI-powered search startup of using its content without permission. This move marks the second time the NYT has confronted a company for allegedly misappropriating its material. According to reports, the Times claims Perplexity is accessing and utilizing its content to generate summaries and other outputs, actions it argues infringe on copyright laws. The startup now has two weeks to respond to the accusations. A Growing Pattern of Tensions Perplexity AI is not the only publisher-facing scrutiny. In June, Forbes threatened legal action against the company, alleging “willful infringement” by using its text and images. In response, Perplexity launched the Perplexity Publishers’ Program, a revenue-sharing initiative that collaborates with publishers like Time, Fortune, and The Texas Tribune. Meanwhile, the NYT remains entangled in a separate lawsuit with OpenAI and its partner Microsoft over alleged misuse of its content. A Strategic Legal Approach The NYT’s decision to issue a cease-and-desist letter instead of pursuing an immediate lawsuit signals a calculated move. “Cease-and-desist approaches are less confrontational, less expensive, and faster,” said Sarah Kreps, a professor at Cornell University. This method also opens the door for negotiation, a pragmatic step given the uncharted legal terrain surrounding generative AI and copyright law. Michael Bennett, a responsible AI expert from Northeastern University, echoed this view, suggesting that the cease-and-desist approach positions the Times to protect its intellectual property while maintaining leverage in ongoing legal battles. If the NYT wins its case against OpenAI, Bennett added, it could compel companies like Perplexity to enter financial agreements for content use. However, if the case doesn’t favor the NYT, the publisher risks losing leverage. The letter also serves as a warning to other AI vendors, signaling the NYT’s determination to safeguard its intellectual property. Perplexity’s Defense: Facts vs. Expression Perplexity AI has countered the NYT’s claims by asserting that its methods adhere to copyright laws. “We aren’t scraping data for building foundation models but rather indexing web pages and surfacing factual content as citations,” the company stated. It emphasized that facts themselves cannot be copyrighted, drawing parallels to how search engines like Google operate. Kreps noted that Perplexity’s approach aligns closely with other AI platforms, which typically index pages to provide factual answers while citing sources. “If Perplexity is culpable, then the entire AI industry could be held accountable,” she said, contrasting Perplexity’s citation-based model with platforms like ChatGPT, which often lack transparency about data sources. The Crux of the Copyright Argument The NYT’s cease-and-desist letter centers on the distinction between facts and the creative expression of facts. While raw facts are not protected under copyright, the NYT claims that its specific interpretation and presentation of those facts are. Vincent Allen, an intellectual property attorney, explained that if Perplexity is scraping data and summarizing articles, it may involve making unauthorized copies of copyrighted content, strengthening the NYT’s claims. “This is a big deal for content providers,” Allen said, “as they want to ensure they’re compensated for their work.” Implications for the AI Industry The outcome of this dispute could set a precedent for how AI platforms handle content generated by publishers. If Perplexity’s practices are deemed infringing, it could reshape the operational models of similar AI vendors. At the heart of the debate is the balance between fostering innovation in AI and protecting intellectual property, a challenge that will likely shape the future of generative AI and its relationship with content creators. Like Related Posts Salesforce OEM AppExchange Expanding its reach beyond CRM, Salesforce.com has launched a new service called AppExchange OEM Edition, aimed at non-CRM service providers. Read more The Salesforce Story In Marc Benioff’s own words How did salesforce.com grow from a start up in a rented apartment into the world’s Read more Salesforce Jigsaw Salesforce.com, a prominent figure in cloud computing, has finalized a deal to acquire Jigsaw, a wiki-style business contact database, for Read more Health Cloud Brings Healthcare Transformation Following swiftly after last week’s successful launch of Financial Services Cloud, Salesforce has announced the second installment in its series Read more

Read More
October 16, 2024in Data, Generative AI, Google, Laws

Subscribe to our mailing list. Join our mail list to receive our newsletter

We'll keep you up to date with our latest news, insights, free resources, and much more.
Don't worry, we won't spam you.
gettectonic.com