2024 Tech Industry Outlook
Google's advances in AI and computing. A LatAm's comparison of policy ideas. AI in the Ports of Latin America and the Caribbean. Recommend books on AI. Synthetic data.
The latest news and trends in the Latin American AI startup scene.
Share your stories to ai@contxto.com.
_
2023: A Year of Groundbreaking Advances in AI and Computing
Google Research and Google DeepMind have made significant advancements in the field of AI throughout 2023. They have focused on developing and delivering innovative and beneficial AI applications while also addressing ethical concerns. Some notable achievements include the launch of Bard, a tool for generating text and creative content, and the development of PaLM 2, a language model that excels at advanced reasoning tasks.
They have also introduced Imagen Editor for interactive image editing, Deep Aligner for language skill improvement, and Lyria for AI music generation. Gemini, their most capable AI model, was launched with various applications across different tasks and industries.
In terms of research, they have made progress in scaling vision transformers, algorithmic prompting for reasoning, and improving machine learning model performance. They have also focused on algorithms, optimization, and privacy, with advancements in areas such as graph prediction, load balancing, clustering, and auction design. (Google DeepMind)
What to do on AI policy and why? A Latin American comparison of policy ideas
The paper titled "What to do on AI policy and why? A Latin American comparison of policy ideas" provides a comparative analysis of the discourses, ideas, and expectations related to AI policies across six Latin American countries. It delves into how these countries articulate their national visions and strategies regarding AI, revealing the complex interplay between international influence, regional aspirations, and national identities.
The study examines the extent to which countries have formulated their AI policies independently or in alignment with international organizations like the OECD. It highlights a spectrum where some countries closely follow international guidelines, while others, like Uruguay, show more independence and critical adoption.
The research categorizes countries based on their self-perception as leaders or adopters of AI innovation. Some countries aspire to regional leadership and global influence, while others focus more on adopting and adapting existing AI technologies. (SSRN)
Quick News
AI in the Ports of Latin America and the Caribbean
It outlines the establishment of an ethical framework for AI governance, set by high authorities from various Latin American and Caribbean countries in collaboration with CAF and UNESCO.
The paper emphasizes the strategic role of ports in integrating AI for improved efficiency, competitiveness, and innovation in trade and logistics, ultimately benefiting small and medium-sized enterprises.
Key points include the creation of an Intergovernmental Working Group on AI to strengthen regional capabilities and foster a multisectoral governance approach. It highlights AI's potential to automate port processes, optimize distribution chains, and transform the entire logistics chain, making it a disruptive technology in traditional port activities. (PORTUS Online)
Costa Rica’s Standing in AI Evaluated by Latin American Index
The evaluation was conducted by Chile's National Centre for Artificial Intelligence (Cenia) and presented by the Economic Commission for Latin America and the Caribbean (Cepal). It examined 12 countries based on factors such as enabling conditions, research, development, adoption, and governance of AI.
Costa Rica is ranked eighth out of 12 countries with a score of 38.97 points, while Chile leads with 73.21 points.
The country shows strengths in devices and data centers but faces challenges in human capital advancement, research and development, and AI adoption in the private sector.
Despite a high percentage of the population having internet access, Costa Rica's download speed is below the regional average, and it has not yet implemented 5G technology. (bnn)
Google’s Bard and Duet AI: A New Era of AI Integration in Latin America
This move is expected to open up new avenues for skill development, job creation, and innovation, contributing to the growth of the digital economy in Latin America.
The introduction of Bard and Duet AI is anticipated to spur innovation and generate new employment opportunities while automating routine tasks to free up human resources for more strategic and creative endeavors.
As Latin America embraces AI, a balanced approach is vital. While leveraging AI for growth and innovation, there must also be a robust framework to ensure ethical use and safeguard against potential misuse. (bnn)
Google Invests $1.2 Billion to Boost Digital Transformation in Latin America
This investment is aimed at enhancing productivity, creativity, and sustainability through a variety of initiatives centered on artificial intelligence and digital technologies.
Google is providing 120,000 scholarships for digital skills courses to cater to a diverse range of individuals, aiming to enhance both professional and personal development.
The launch of Google Wallet in various Latin American countries is a significant move towards the digitalization of financial services and promoting financial inclusion.
In Colombia, Google is working on projects to develop digital skills among Colombians, including an agreement with the National Learning Service (Sena) to train instructors and reach tens of thousands of learners. (bnn)
Rampage Is Brazil’s First Vehicle To Use ChatGPT For Customer Service
Ram, the automobile brand, has seen great success in Brazil with the launch of its compact pickup, the Ram Rampage. To enhance customer service, the brand is using AI in the form of ChatGPT on its official WhatsApp channel.
This allows customers to ask questions about the vehicle, from technical inquiries to general information about the product and brand.
The AI system consults the vehicle's owner's manual to provide personalized and practical responses.
The use of AI has resulted in increased customer satisfaction and resolution of queries in the first transaction. (Mopar Insiders)
Anthropic/AI: Unique Structure Amid Safety-First Approach
Led by former OpenAI employees Dario and Daniela Amodei, the company has established a unique governance structure with a "Long-Term Benefit Trust" to prevent direct investor influence.
Anthropic's focus on ethical AI development is demonstrated by its research into bias in its chatbot.
The Long-Term Benefit Trust may attract scrutiny but also positions Anthropic as a model for responsible AI development. As AI continues to evolve, Anthropic's structure and safety-first approach will keep it in the spotlight and fuel ongoing debates about governance and ethics in the AI industry. (Contxto)
OpenAI is in early discussions to raise new funding at a valuation of US$100 billion or more
If successful, this would make OpenAI the second most valuable startup in the US, behind SpaceX. The company is also set to complete a separate tender offer in January, allowing employees to sell shares at a valuation of US$86 billion.
OpenAI's increasing valuation reflects the growing interest in AI following the success of its chatbot, ChatGPT.
The company has also been in talks to raise funds for a new chip company. OpenAI recently faced leadership turmoil but has since refocused on its products. (Contxto)
US AI Risk Agency Struggles with Funding for Critical Safety Standards
The National Institute of Standards and Technology (NIST) is facing challenges in developing robust safety standards for AI by July 2024.
The agency's current budget is inadequate, leading to concerns about over-reliance on the private sector.
Recent discussions at the NeurIPS AI conference highlighted the disparity between NIST's resources and those of private AI developers.
The secretive nature of commercial AI models further complicates the standardization process. Experts argue for NIST's role as a neutral body and stress the need for substantial support to fulfill its mandate effectively. (Contxto)
VCs Show Unusual Interest in AI Giants OpenAI and Anthropic
VC firm Menlo Ventures is leading a significant investment round in AI company Anthropic, valuing it at over US$15 billion. This marks a shift in the VC landscape, as early-stage investors are now putting money into mature AI companies.
Other investors, like Arlan Hamilton from Backstage Capital, are also showing interest in prestigious AI investments.
However, there are concerns about the high valuations and uncertainties of investing in high-growth, capital-intensive tech sectors.
The appetite for AI investments is growing, indicating a belief in the long-term potential of AI technology. (Contxto)
Apple Focuses on Running AI Directly on iPhones, Diverging from Cloud Use
Apple is researching to run large language models (LLMs) on smartphones, moving away from the cloud-based approach. Their paper titled "LLM in a Flash" demonstrates a solution for running LLMs effectively on devices with limited memory.
This research aims to address the computational challenges of operating AI models on mobile devices. Unlike competitors, Apple is focusing on device-based AI to enhance data security and privacy.
This shift in strategy could redefine the smartphone experience by offering offline-capable AI assistance and advanced privacy measures.
While not indicating specific product plans, this initiative sets a precedent for integrating AI into various devices and applications. (Contxto)
AI Education
Outfit Anyone: Ultra-high quality virtual try-on for Any Clothing and Any Person
The article introduces "Outfit Anyone," a virtual try-on system that allows users to try on clothing virtually without physically wearing them. The system uses a two-stream conditional diffusion model that handles garment deformation for realistic results. It can work with various scenarios, from realistic clothing changes to eccentric fashion styles and even anime characters. The system also accommodates different body shapes and includes a post-hoc refiner to enhance the texture and realism of the clothing. Additionally, it can be integrated with a pose-to-video model for outfit changes and motion video generation.
Database Fundamentals
Understanding the fundamental problems in databases, such as durability, atomicity, isolation, and performance, is crucial for database engineers. Storage engines play a critical role in database performance by providing an abstraction for reading and writing data to persistent storage, to minimize disk I/O and seeks.
ChatGPT prompt engineering for beginners in 2024 – Full Guide
Prompt engineering is a crucial skill for effectively interacting with advanced AI tools like ChatGPT. It involves crafting requests in a way that yields accurate and useful responses. To optimize your interactions with ChatGPT, it is important to establish clear objectives, provide relevant context, and be precise in your instructions.
Food for Thought
Top 26 Books On AI As Gift
This article discusses some of the best books on AI and the lessons they teach. It highlights a range of books that cover various topics in AI, including machine learning, robotics, and natural language processing.
The article also explores books that focus on the practical applications of AI in business and finance. Additionally, it mentions philosophical books that delve into the ethical implications and potential risks of AI.
The article concludes by recommending books that examine the history and pioneers of AI, as well as those that discuss the limitations and challenges of the field. (Artificial intelligence +)
From AI to Sustainability: 2024 Tech Industry Outlook
This article discusses key trends shaping major industries in the coming years, including AI, cloud computing, blockchain, electric vehicles, and more. It highlights the growth of generative AI and the challenges it faces in gaining public trust.
The semiconductor industry is expected to experience an upswing, driven by demand for AI chips. AI governance and regulation are evolving slowly, with different approaches in different regions. The article also explores trends in cybersecurity, agriculture, blockchain adoption, and the growth of cloud computing and edge computing.
It discusses the pressure on natural resource supplies due to the transition to electric vehicles and renewable energy. Additionally, it touches on the expected return of cryptocurrency markets to a bull cycle and the potential approval of spot-based ETFs for Bitcoin and Ethereum. (Curious Compass)
What is Synthetic Data and How to Work with it?
Synthetic data is artificially generated data that mimics real-world data and is used for research, training machine learning algorithms, data analysis, and testing software products. It was originally used to conceal personally identifiable information (PII) within datasets.
The demand for synthetic data has increased due to the growing popularity of large language models (LLMs) that require a huge volume of data for training. Synthetic data can help address issues such as bias in datasets, reduce cost and time in testing, enable thorough testing and validation of models, facilitate innovation and experimentation, and ensure compliance with privacy regulations.
There are various methods to generate synthetic data depending on the task at hand. The synthetic data market is projected to grow significantly in the coming years. (Turing Post)
_
Jobs
Lead Data Scientist, Data Science (GTM Advanced Analytics) - HP (Tlaquepaque, Jalisco, Mexico; Hybrid)
Senior Data Scientist - HP (Tlaquepaque, Jalisco, Mexico; Hybrid)
Principal Data Scientist - Grupo Modelo (Mexico City Metropolitan Area; On-site)