Artificial General Intelligence under our nose: Unpacking the skepticism surrounding modern AI
Artificial General Intelligence (AGI) is here. Notes from the AI Engineer Summit. Two freshly reports on the status of AI in the last 12 months. LLMs for Dummies.
The latest news and trends in the Latin American AI startup scene.
Share your stories to ai@contxto.com.
_
Artificial General Intelligence Is Already Here
The article discusses the concept of Artificial General Intelligence (AGI) and argues that the current generation of advanced AI large language models already exhibit important aspects of AGI.
These "frontier models" can perform a wide variety of tasks, operate on different modalities, handle multiple languages, and learn from prompts. However, there is reluctance to acknowledge AGI due to skepticism about metrics, ideological commitments, concerns about human exceptionalism, and economic implications.
There is a reluctance to acknowledge artificial general intelligence due to skepticism about metrics, alternative AI theories, human exceptionalism, and concerns about the economic implications. (Noema Magazine).
The State of AI Engineering
The inaugural AI Engineer Summit took place with over 500 attendees and featured major announcements such as new coding models from Replit, GitHub's USD$100 million revenue from its AI code completion tool, and AutoGPT's USD$12 million investment.
The conference explored the challenges and possibilities in AI engineering, highlighting the importance of prompt management, evaluations, user experience, and guardrails to ensure reliable AI systems.
The conference emphasized that even small AI projects can have significant impacts. The idea of a "1000x AI engineer" was discussed, envisioning software engineers enhanced by AI tools, building AI products, or even being replaced by AI.(Artificial Ignorance)
Quick News
AI stands out in new batch from Platanus Ventures
The upcoming Demo Day by Platanus Ventures will feature startups that have incorporated Artificial Intelligence (AI) as a central element in their products.
AI has also been instrumental in the selection process for startups at Platanus Ventures since 2022. The event will showcase startups from various industries such as SaaS, foodtech, and petcare.
The Demo Day is an exclusive online event for selected investors and will take place from October 24 to October 31. (Contxto)
MIT recognizes young Peruvians for their advancements in Artificial Intelligence and robotics
Peru has emerged as the country with the most award recipients in the Innovators Under 35 LATAM 2023 (IU35 LATAM) awards, which celebrate young Latin Americans using science and technology to tackle regional challenges.
Eight Peruvian innovators were honored for their work in robotics and Artificial Intelligence. Their projects range from sustainable water purification to sign language translation using AI.
The awards ceremony will be hosted in Peru, highlighting the country's growing tech sector. Brazil, Chile, and Mexico are also making strides in innovation, while Peru has seen a drop in its global ranking. (Contxto)
Wizeline and Tec de Monterrey drive the first Generative Artificial Intelligence Laboratory in LatAm
The Tecnológico de Monterrey and Wizeline have joined forces to establish the first Generative Artificial Intelligence Laboratory in Mexico and Latin America.
The lab, called G.AI.L, will be located on the Guadalajara campus of Tec de Monterrey and will serve as a hub for academic research, entrepreneurship, and practical applications of AI.
With a significant investment from Wizeline, G.AI.L aims to become the leading center for Generative AI in Mexico, fostering collaboration between researchers, students, and businesses. The lab will offer a program that combines theory and practice, including internships, hackathons, and conferences. (Contxto)
Puscar Buritica Family Office: The foundation helping to deliver economic change
Latin America has made significant progress in technology adoption and access to digital services in recent years. However, access to economic opportunities remains a challenge, with a large portion of the population living in poverty and lacking basic financial services.
The Puscar Buriticá Family Office, founded by Michael Puscar and Johana Buritica, aims to address these issues by promoting digital inclusion and providing opportunities in technology for children.
The family office also supports the development of smart cities, which can improve the lives of citizens through innovation and increased efficiency. Additionally, the office seeks to support local entrepreneurs and tech talent by providing mentorship and expertise. (The Sociable)
Millions of Workers are training AI Models for Pennies
The global data collection and labeling market, valued at $2.22 billion in 2022, is expected to grow to $17.1 billion by 2030.
Crowdsourcing platforms like Appen, which provides data labeling services for companies like Amazon and Google, rely on workers in developing countries who are paid low wages.
These workers, often in Venezuela, India, and the Philippines, are struggling to make a living, with irregular work and low pay. Some experts see this as a form of data colonialism, where workers in the Global South label data that will be used in the Global North. (Wired)
Brazil Consults On Regulatory Sandbox For AI And Data Protection
The Brazilian National Data Protection Authority (ANPD) has launched a public consultation for its pilot regulatory sandbox program focused on testing AI technologies while adhering to data protection rules.
The sandbox aims to facilitate collaborative AI experimentation within a regulatory framework and promote responsible innovation.
The ANPD hopes to gather input from stakeholders to shape future regulations and promote transparency and inclusive decision-making. The sandbox exercise is supported by a technical study and collaboration with the Development Bank of Latin America and the Caribbean. (Forbes)
Researchers Measure Global Consensus over the Ethical Use of AI
To understand the global consensus on AI ethics, a team of researchers from Brazil conducted a systematic review and meta-analysis of global guidelines for AI use.
Their findings, published in the journal Patterns, revealed that while many guidelines emphasized the importance of privacy, transparency, and accountability, fewer guidelines valued truthfulness, intellectual property, or children’s rights.
Additionally, most guidelines outlined ethical principles and values but did not offer practical methods for implementation or advocate for legally binding regulations. (Lab Manager)
B3 invests USD$1.9 million in deep tech MBOCHIP
The Sao Paulo Stock Exchange, B3, has acquired a minority stake in deep-tech startup MBOCHIP. The initial investment is approximately USD$1.9 million, although the exact stake percentage is undisclosed.
B3 aims to provide its clients with new functionalities and improved services through this partnership. MBOCHIP, an advanced technology company, will continue to operate independently and expand its customer base.
The investment aligns with B3's modernization strategy, which includes previous investments in Vermiculus and collaboration with Nasdaq. Deep tech companies offer innovative solutions through scientific or technological advancements and present attractive investment opportunities. (Contxto)
LATAM startups selected for CIC International Soft Landing
The Cambridge Innovation Center (CIC) and ProChile have announced the eight health tech startups selected for the CIC International Soft Landing program.
The program aims to support these startups in expanding their businesses in the United States by providing them with a six-week online training. The training will cover topics such as the healthcare technology sector in the US, differences between the Chilean and American markets, mentorship, and regulatory compliance insights.
The selected startups include companies focused on clinical simulators, medical inventory management, custom craniofacial implants, innovative sports products, hospital management platforms, clinical management software, modular construction systems, and first aid products. (Contxto)
Mexico’s Data-Driven Transformation and AI Warfare
Only 54% of Mexican companies are effectively utilizing their data resources, highlighting a need for streamlined data solutions.
Mexico's nearshoring potential depends on financial data and infrastructure development. AI chatbots are expected to revolutionize customer engagement, while Microsoft has unveiled healthcare data and AI tools to streamline access to information. (Mexico Business News)
AI Education
Introduction to Generative AI
This is an introductory-level microlearning course aimed at explaining what Generative AI is, how it is used, and how it differs from traditional machine learning methods. It also covers Google Tools to help you develop your own Gen AI apps.
Maestría en Inteligencia Artificial
The Master's in Artificial Intelligence (MaIA) at the University of Los Andes is designed for Spanish-speaking professionals with basic knowledge in programming, linear algebra, probability, and statistics.
The program covers various AI methods and applications such as machine learning, deep learning, natural language processing, computer vision, semantic web, and autonomous systems.
Deep learning course
The course covers various topics in deep learning, including machine learning fundamentals, multi-layer perceptrons, convolutional layers, optimization techniques, autoencoders, computer vision, and more. The course materials include slides, recordings, and a virtual machine.
Food for Thought
State of AI Report 2023
Published by AI investors Nathan Benaic and the Air Street Capital team, the State of AI Report 2023 analyzes the latest developments in AI and aims to spark informed discussions about its current state and future implications.
The report covers various dimensions including research, industry, politics, safety, and predictions.
Key themes in the 2023 report include the dominance of GPT-4, advancements in LLMs and diffusion models in the life sciences, the growing importance of compute power, the success of AI startups focused on generative AI applications, the increasing focus on AI safety, and challenges in evaluating state-of-the-art models. (State of AI)
Kaggle AI Report 2023
The 2023 Kaggle AI Report was created by Kaggle’s community and selected from hundreds of submissions. Each paper within this report gives a unique viewpoint on the most interesting or most important recent development in the field of AI and ML.
The report is broken down into the seven sections listed below:
Generative AI
Text data
Image / video data
Tabular / time series data
Kaggle competitions
AI ethics
Other (Kaggle)
LLMs for Dummies
Large language models (LLMs) have improved exponentially in recent years, with GPT-3 having 175 billion parameters and GPT-4 rumored to have even more, resulting in significant advancements in natural language processing.
LLMs represent words with numbers called word vectors, which are stored in clusters based on their relationships. The models are trained on vast amounts of language data to improve their understanding of context and meaning.
Transformer models, which consist of multiple layers, help the LLMs grasp the nuances of language. The article also mentions the potential biases inherited by LLMs due to their training data. (Digital Native)
_
Jobs
Senior Python - Viajes El Corte Inglés México (Mexico City, hybrid)
Full Stack Sr. Developer - Keasis (Mexico City, remote)
Programador React Senior - Turing (Mexico, remote)