Goldman Sachs investment bank published a 31-page report titled "Gen AI: Too Much Spend, Too Little Benefit?" wonders if this technology is a bubble and states that this is "unreliable, unsustainable, requires an entire rebuild of America's power grid, and is most decidedly not the future." "Tech giants and beyond are set to spend over $1tn on AI capex in coming years, with so far little to show for it. So, will this large spend ever pay off?" questions Goldman Sachs. The investment bank states, "The technology isn’t designed to solve the complex problems that would justify the costs, which may not decline as many expect." "Even if it does, we explore whether the current chips shortage and looming power shortage will constrain AI growth." Goldman Sachs mentions that utility companies will have to spend nearly 40% more in the next three years to keep up with the demand from hyperscalers like Google and Microsoft. It also adds that only about 5% of companies report using generative AI in regular production. The report includes an interview with economist Daron Acemoglu of MIT, an Institute Professor who published a paperback in May called "The Simple Macroeconomics of AI" that argued that "the upside to US productivity and, consequently, GDP growth from generative AI will likely prove much more limited than many forecasters expect." Acemoglu declared that "truly transformative changes won't happen quickly, and few—if any—will likely occur within the next 10 years." "AI's ability to affect global productivity is low because "many of the tasks that humans currently perform are multi-faceted and require real-world interaction, which AI won't be able to improve anytime soon materially." Newsletter: Goldman Sachs has called BS on Generative AI, and I believe that it's time that everybody follows suit – generative AI is unreliable, unsustainable, requires an entire rebuild of America's power grid, and is most decidedly not the future.https://t.co/YULEkHYBFP — Ed Zitron (@edzitron) July 8, 2024
Amazon Web Services (AWS) announced several product developments during its annual summit in New York, including an App Studio that enables enterprise non-technical users to build internal applications with simple prompts. Available under a limited preview, AWS App Studio "opens up application development to an entirely new set of builders," said Dilip Kumar, vice president of applications at AWS. The company claims App Studio is better than most low-code development tools, which do the job but fail to produce fully secure apps that comply with enterprise privacy and security policies. It competes with other app creation platforms such as Creatio’s Quantum and Salesforce Platform. Currently, enterprise operations depend on data-heavy internal processes and workflows. Announcing @awscloud App Studio, the fastest and easiest way to build applications! Start with a prompt, and end with a real AWS application – fully managed by AWS. It’s been an amazing journey and working on this! pic.twitter.com/lhIC12Y7dN — Adam Seligman (@adamse) July 10, 2024
Microsoft Research announced this month that it open-sourced GraphRAG, a valuable Python library that extracts insights from text. It uses LLMs to automate the extraction of knowledge graphs from datasets and text documents. "GraphRAG is a structured, hierarchical approach to Retrieval Augmented Generation (RAG), as opposed to naive semantic-search approaches using plain text snippets," explained the company in a blog post. Retrieval-augmented generation (RAG) is a technique to improve LLM outputs using real-world information. RAG searches for information based on a user query on private datasets, data that the LLM is not trained on and has never seen before, such as an enterprise’s proprietary research, business documents, or communications. Microsoft just open-sourced GraphRAG. It might be the best Python library to extract insights from text. Much more powerful than vanilla RAG. It uses LLMs to automate the extraction of knowledge graphs from your datasets and text documents. !pip install graphrag pic.twitter.com/BhVrPkhZCs — Lior⚡ (@AlphaSignalAI) July 9, 2024
Poe, Quora's subscription-based cross-platform aggregator of chatbots, introduced yesterday a feature called Previews that lets users build interactive web experiences directly in chats. It requires paying $20 per month for Poe’s premium plan. Previews works particularly well with LLMs that excel at coding, including Claude 3.5 Sonnet, GPT-4o, and Gemini 1.5 Pro. These applications include games, animations, drag-and-drop interfaces, and data visualizations, and support HTML output with CSS and Javascript functionality. Users can also leverage Poe features like multi-bot chat, file upload, and video input to help you build your custom web applications. They can be shared using a dedicated link. These are some examples: Generate an interactive presentation from a financial report Create dynamic Flashcards around a topic Develop an interactive Drum Machine Collide colors to form new shades Poe said it plans to introduce support for additional formats in the coming days and weeks. Previews are like Anthropic’s recently introduced Artifacts: dedicated workspaces where users can edit and add to AI-generated content like code and documents.
Cornell University launched an innovative online certificate program, Designing and Building AI Solutions, authored by Cornell Senior Visiting Lecturer Lutz Finger, who generated an AI clone of himself to update the courses with new content continuously. Lutz Finger explained that his AI clone is supercharging him to enhance his teaching capabilities. "In the courses, I use him as a case study to identify all of the product, business, and design decisions that need to be considered," he said. This certificate program from the Cornell SC Johnson College of Business also includes an AI-powered co-pilot, available 24/7, that provides coding assistance to help non-technical professionals build products. Available through eCornell, the program provides a hands-on deep dive into prompt engineering and machine learning applications and data handling, ethics, and compliance. Five online courses take learners on a virtual journey to multiple locations worldwide to examine real-life applications for AI products across industries. Finger’s style of integrated lectures provides edutainment as he visits ski slopes, cow pastures, airport terminals, and more. "I designed this program to be entertaining while also breaking new ground by weaving the AI tools we are learning about into every facet of the program—from actual course delivery to hands-on practice," said Finger. The program is designed for managers, leaders, innovators, and anyone interested in applying AI and automation in their industry. Participants will be placed in small, virtual classes with peers to maximize learning and support networking and collaboration.
