Jagged Intelligence Archives - gettectonic.com
Salesforce's Enterprise General Intelligence
20Jun

Salesforce’s Enterprise General Intelligence

Salesforce is carving a distinct path in the AI landscape, diverging from the industry’s pursuit of Artificial General Intelligence (AGI). Instead, the company is tackling a pressing, practical challenge: ensuring AI is reliable for enterprise use. Salesforce’s Enterprise General Intelligence (EGI) framework prioritizes consistency, safety, and trustworthiness over speculative potential, aiming to deliver dependable AI for real-world business applications. The EGI FrameworkLarge language models (LLMs) excel at tasks like drafting emails or analyzing datasets but often exhibit “jagged intelligence”—impressive in some areas yet prone to basic errors or fabrications, known as hallucinations. These inconsistencies pose significant risks in enterprise settings, where errors can lead to compliance issues, financial losses, or eroded customer trust. Salesforce’s EGI framework addresses this by focusing on infrastructure that ensures AI reliability today, rather than chasing futuristic goals. From Inconsistency to DependabilitySalesforce likens LLMs to “an intern who writes flawless code but forgets to save the file.” To address this uneven performance, the company is enhancing its AI agents with layered reinforcement to boost consistency. Central to this effort is Agentforce, Salesforce’s agentic system, supported by the Atlas Reasoning Engine, which integrates internal and external data for more accurate reasoning and retrieval. Together, these form the core of EGI, aiming to make digital labor predictable and trustworthy. Rigorous Testing in Real-World ScenariosRather than relying solely on traditional benchmarks, Salesforce introduced CRMArena, a simulated environment that tests AI agents on practical CRM tasks like service support and analytics. Initial results show success rates below 65%, even with guided prompting, underscoring the challenges. However, this is precisely Salesforce’s point: stress-testing AI in realistic conditions exposes weaknesses before deployment, ensuring reliability in customer-facing roles. A Platform for Enterprise TrustSalesforce emphasizes that enterprises need more than powerful models—they require systems guaranteeing predictability and accountability at scale. EGI is positioned as a practical, present-focused solution, sidestepping AGI hype to deliver AI that businesses can trust today. While its long-term impact remains to be seen, Salesforce’s approach signals a pragmatic step toward reliable, enterprise-ready AI. Like Related Posts Who is Salesforce? Who is Salesforce? Here is their story in their own words. From our inception, we’ve proudly embraced the identity of Read more Salesforce Marketing Cloud Transactional Emails Salesforce Marketing Cloud Transactional Emails are immediate, automated, non-promotional messages crucial to business operations and customer satisfaction, such as order Read more Salesforce Unites Einstein Analytics with Financial CRM Salesforce has unveiled a comprehensive analytics solution tailored for wealth managers, home office professionals, and retail bankers, merging its Financial Read more AI-Driven Propensity Scores AI plays a crucial role in propensity score estimation as it can discern underlying patterns between treatments and confounding variables Read more

Read More
June 20, 2025in Data, Enterprise, Salesforce
The Paradox of Jagged Intelligence in AI
20May

The Paradox of Jagged Intelligence in AI

AI systems are breaking records on complex benchmarks, yet they falter on simpler tasks humans handle intuitively—a phenomenon dubbed jagged intelligence. This ainsight explores this uneven capability, tracing its evolution in frontier models and the impact of reasoning models. We introduce SIMPLE, a new public benchmark with easy reasoning tasks solvable by high schoolers, vital for enterprise AI where reliability trumps advanced math skills. Since ChatGPT’s 2022 debut, foundation models have been marketed as chat interfaces. Now, reasoning models like OpenAI’s o3 and DeepSeek’s R1 leverage extra inference-time computation for step-by-step internal reasoning, boosting performance in math, engineering, and coding. This shift to scaling inference compute arrives as pretraining gains may be plateauing. Benchmarking the Gaps Traditional AI benchmarks measure peak performance on tough tasks, like graduate exams or complex code, creating new challenges as old ones are mastered. However, they overlook reliability and worst-case performance on basic tasks, masking jaggedness in “solved” areas. Modern models outshine humans on some challenges but stumble unpredictably on others, unlike specialized tools (e.g., calculators or photo editors). Despite advances in modeling and training, this inconsistent jaggedness persists. SIMPLE targets easy problems where AI still lags, offering insights into jaggedness trends. Evolution of Jaggedness Will jaggedness shrink or grow as models advance? This question shapes enterprise AI success. Lacking jaggedness benchmarks, we created SIMPLE—a dataset of 225 simple questions, each solvable by at least 10% of high schoolers. Example Questions from SIMPLE Performance Trends Evaluating current and past top models on SIMPLE traces jaggedness over time. Green tasks are high school-level; blue are expert-level. School-level benchmarks saturated by 2023-2024, shifting focus to harder tasks. SIMPLE, using the best of gpt-4, gpt-4-turbo, gpt-4o, o1, and o3-mini, scores lowest on school-level questions. Yet, reasoning models show a ~30% improvement, suggesting they reduce jaggedness by double-checking work, linking reasoning to better simple-task performance. Case Study Insights and Implications Reasoning models transfer top-line gains to simple tasks to some extent, but SIMPLE remains unsaturated. Jaggedness persists, with top-line progress outpacing worst-case improvements. This mirrors computing’s history: excelling in narrow domains, outpacing human limits once applied, yet always facing new challenges. Jaggedness may not just define AI—it could be computation’s inherent nature. Like Related Posts Who is Salesforce? Who is Salesforce? Here is their story in their own words. From our inception, we’ve proudly embraced the identity of Read more Salesforce Unites Einstein Analytics with Financial CRM Salesforce has unveiled a comprehensive analytics solution tailored for wealth managers, home office professionals, and retail bankers, merging its Financial Read more AI-Driven Propensity Scores AI plays a crucial role in propensity score estimation as it can discern underlying patterns between treatments and confounding variables Read more Tectonic’s Successful Salesforce Track Record Salesforce Technology Services Integrator – Tectonic has successfully delivered Salesforce in a variety of industries including Public Sector, Hospitality, Manufacturing, Read more

Read More
May 20, 2025in Data, Enterprise
08May

Salesforce Tackles Enterprise AI Reliability with Enterprise General Intelligence (EGI)

As businesses increasingly adopt AI, a critical challenge has emerged: inconsistent performance in real-world applications. Salesforce calls this phenomenon “jagged intelligence”—where AI excels in controlled environments but falters under dynamic enterprise demands. To address this, Salesforce is pioneering Enterprise General Intelligence (EGI), a new framework designed to ensure AI is not just powerful but reliable, consistent, and safe for business use. Why Enterprise AI Needs a New Approach Traditional AI benchmarks often fail to reflect real-world enterprise needs. Issues like: …have made many companies hesitant to fully deploy AI at scale. Salesforce’s EGI rethinks AI alignment for enterprises, prioritizing:✔ Consistency – Reliable performance across diverse business cases✔ Specialization – Task-specific AI models over generic LLMs✔ Safety & Governance – Built-in guardrails for compliance Key Innovations Powering EGI 1. SIMPLE: Measuring AI Consistency Salesforce’s SIMPLE dataset (225 reasoning questions) evaluates how AI performs under varying conditions—helping identify and fix inconsistencies before deployment. 2. CRMArena: Real-World AI Testing This benchmarking framework simulates authentic CRM scenarios (service agents, analysts, managers) to ensure AI adapts to real business needs—not just lab conditions. 3. SFR-Embedding: Smarter Enterprise AI A new embedding model (ranked #1 on MTEB’s 56-dataset benchmark) enhances AI’s ability to understand complex business data, improving decision-making in Salesforce Data Cloud. 4. xLAM V2: AI That Takes Action Unlike text-only LLMs, Large Action Models (xLAM V2) predict and execute enterprise tasks—optimizing everything from inventory management to financial forecasting with high precision. 5. SFR-Guard & ContextualJudgeBench: AI Safety Co-Innovation: Doubling AI Accuracy with Customer Feedback Salesforce’s customer-driven development has already doubled AI accuracy in key applications. Itai Asseo, Senior Director of Incubation & Brand Strategy at Salesforce: “By working directly with enterprises, we’ve refined AI to outperform competitors in real-world use cases—boosting both performance and trust.” The Future of Enterprise AI Salesforce’s EGI framework is setting a new standard: AI that works as reliably in business as it does in theory. For telecom and tech leaders, this means:✅ Fewer AI surprises – Consistent, predictable outputs✅ Higher ROI – Specialized models for key workflows✅ Stronger compliance – Built-in governance & safety As AI evolves, Salesforce is ensuring enterprises don’t just adopt AI—they can depend on it. Next Steps: The era of reliable enterprise AI is here. Like Related Posts Who is Salesforce? Who is Salesforce? Here is their story in their own words. From our inception, we’ve proudly embraced the identity of Read more Salesforce Marketing Cloud Transactional Emails Salesforce Marketing Cloud Transactional Emails are immediate, automated, non-promotional messages crucial to business operations and customer satisfaction, such as order Read more Salesforce Unites Einstein Analytics with Financial CRM Salesforce has unveiled a comprehensive analytics solution tailored for wealth managers, home office professionals, and retail bankers, merging its Financial Read more AI-Driven Propensity Scores AI plays a crucial role in propensity score estimation as it can discern underlying patterns between treatments and confounding variables Read more

Read More
May 8, 2025in Data, Enterprise, Salesforce, Salesforce Data Cloud

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