We are so back - August 29, 2024

Dear all,

These are not the best times for people involved in software development. Layoffs, uncertainty about return on investment in software, budget cuts, especially in innovative projects that carry risk but may bring future value.?

The last few quarters have not been easy for me either, which explains why I put this? newsletter on hold.?

However, I realized that what gives me true joy is sharing things with people. Some of you asked why I stopped doing it.?

I don’t know.

No matter the current landscape, the time for innovators is still exciting – tools like low-code allow for rapid prototyping, and AI enables faster access to relevant issues by processing our thoughts. I believe that the progress between user technologies, software, and hard science, art, design, or culture is still very visible.?

This continues to excite me greatly, which is why I decided to return to writing about these things.

What I'll be sharing, as in the past, are simply things that caught my attention. Things that I find interesting.. I'll focus less on hard business and more on things that may not be obvious but could be inspiring to you. Let me know if you dig this format.

We're back, baby.?

–

WhatsApp: The Unlikely E-commerce Giant of Latin America

?? WhatsApp has become a dominant e-commerce platform in Latin American countries like Brazil and Chile. During the pandemic, it was the preferred choice for online shopping in the region. However, WhatsApp wasn't designed for e-commerce, leading to challenges for businesses. Enter Mercately, a startup building back-end software to help brands sell more effectively through WhatsApp.

? Why it matters: This trend highlights the importance of adapting to local consumer behaviors rather than imposing external models. It also demonstrates the potential for innovative solutions that bridge the gap between existing popular platforms and business needs.

?? My take: The success of WhatsApp as an e-commerce platform in Latin America echoes the rise of super-apps in Asia. It's a reminder that innovation doesn't always mean creating new platforms, but sometimes leveraging existing ones in creative ways.

?? Click for more information

The Surprising Limitations of AI: Why Large Language Models Can't Count

?? Despite their ability to process vast amounts of data and perform complex tasks, large language models (LLMs) like GPT-4 and Claude sometimes fail at seemingly simple tasks, such as counting the number of times a letter appears in a word. This limitation stems from the fundamental architecture of these models, which process text as tokens rather than individual letters.

? Why it matters: Understanding the limitations of AI is crucial as we increasingly rely on these technologies. It reminds us that AI, despite its impressive capabilities, processes information very differently from humans and has distinct blind spots.

?? My take: These limitations serve as a reality check on the current state of AI. While AI has made remarkable progress, it's important to remember that these systems are not infallible or omniscient. This understanding can help us design more effective AI systems and set realistic expectations for their use in various applications.

?? Click for more information

Modernizing a 142-Year-Old Company: PPG's Digital Transformation Journey

?? PPG, a 142-year-old company, is undergoing a significant digital transformation to modernize its operations and reduce technical debt. The company is focusing on three key areas: customer-facing technologies, back-office systems, and technical architecture. They're adopting a cloud-only strategy and developing innovative digital solutions to drive revenue and improve efficiency.

? Why it matters: This case study demonstrates how even long-established companies can embrace digital transformation to stay competitive. It shows the importance of balancing legacy pride with the need for modernization.

?? My take: PPG's approach to transformation, focusing on both customer-facing and back-end systems, is commendable. Their strategy of embedding IT teams with business units to intuitively understand needs is particularly noteworthy. This approach could serve as a model for other legacy companies looking to modernize. Also shows innovation can be (should be?) an iteration instead of revolution.?

?? Click for more information

DNA Computers: A Leap Forward in Data Storage and Processing

?? Researchers have developed a new technology that could store petabytes of data in DNA for thousands or even millions of years. The system, which uses a soft polymer material as a scaffold for DNA, can also process data, as demonstrated by solving sudoku puzzles.

? Why it matters: This breakthrough could revolutionize data storage, offering unprecedented capacity and longevity. It also brings us a step closer to realizing DNA computers, which could offer unique advantages in certain computational tasks.

?? My take: While this technology sounds like science fiction, it represents a fascinating convergence of biology and computing. The potential for long-term, high-density data storage is particularly intriguing, especially as we grapple with ever-increasing amounts of data.

?? Click for more information

AI Governance: Open Source Exemption in Kill Switch Legislation

?? A recent Reddit discussion highlighted an important update to proposed AI legislation. The bill now includes an exemption for open-source AI models from the "kill switch" requirement. The relevant text specifies that the full shutdown applies to models "controlled by a developer," effectively exempting open-source AIs from this particular regulation.

? Why it matters: This exemption addresses concerns that mandatory kill switches would effectively end open-source AI development. It demonstrates legislators' attempts to balance safety measures with the benefits of open-source innovation in AI.

?? My take: This legislative update shows the complexity of regulating AI development. While safety measures like kill switches are crucial for large, controlled AI systems, preserving the open-source ecosystem is equally important for fostering innovation and democratizing AI technology. The challenge lies in striking the right balance between these competing interests.

?? Click for more information

The Generational Shift in Communication: The Decline of Phone Calls

?? A recent survey found that a quarter of people aged 18 to 34 never answer phone calls, preferring text messages or other forms of digital communication. This trend is reshaping both personal and professional communication norms.

? Why it matters: Understanding changing communication preferences is crucial for businesses looking to engage with younger demographics. It also has implications for workplace communication and customer service strategies.

?? My take: This shift in communication preferences reflects broader changes in how we manage our time and attention. While there's value in direct verbal communication, the asynchronous nature of text-based communication offers its own advantages. Businesses and individuals will need to find the right balance to maintain effective communication across generations.

?? Click for more information

Quote of the week: by Ido Lustig

Eye-popping number: 10 PB per cm3 – the data storage density achieved by the new DNA storage technology, equivalent to 10 million GB in a space the size of a sugar cube.

Twitter Banger You Need to Retweet:

That's all for this week's Innovation Brief. As always, I encourage you to share this newsletter with friends and colleagues who might find it valuable. Your feedback is crucial in shaping future editions, so please don't hesitate to reach out with your thoughts and suggestions.

Until next time, stay curious and keep learning.

Radek Zaleski

P.S. If you enjoyed this edition, please consider sharing it with your network. Or replying to me – it really gets me going.