As the second quarter of 2024 comes to a close, I wanted to look back at some of the many highlights and developments in the world of data, AI, and technology as a whole. As I have said many times in the past, technology moves at such a fast pace that if you blink, you will be sure to have missed something. While some of the things I cover are too big to miss, I want to make sure that I capture the things that I find the most fascinating and relevant to today as well as share my thoughts on where things are headed. Today, we will be looking at a collection of 5 topics, stories, and releases that I hope you find as interesting as I do.
We will be looking at a collection of 5 topics, stories, and releases that I hope you find as interesting as I do.
Reading the report criteria, they give almost equal weighting of categories across the board except for four that are worth twice as much. These are Data Storage/Formats, End-to-End Integration, Deployment Options, and Generative AI/LLM. I have a bone to pick with this. I understand the value that generative AI brings, but truth be told, it is not involved with every use case in a data lakehouse. You know what is involved? Security/governance, data quality, data transformation, and performance optimization—all categories worth less than generative AI. At least make the title of the report “Data Lakehouses for AI” or something because, let’s be honest, for every 1 person that reads the actual report, 9 others just see the diagram on LinkedIn.
With that out of the way, I am not surprised to see Databricks, Google, and Snowflake in the leaders category. Having worked hands-on with all of the major cloud data lakehouses, those three definitely felt the most mature in their offering while maintaining ease of use. I am surprised to see AWS and Microsoft fall short to Salesforce, Oracle, Teradata, and Cloudera in terms of current offerings. And I thought for sure that Dremio would have made even a small appearance with their presence in the space.
As always, I recommend that you read the report to understand the full evaluation criteria, especially if you are using it as part of your company’s evaluation.
Cindi Howson’s (ThoughtSpot) podcast, The Data Chief, has become a well-known platform for insightful, inspiring conversations and interviews with some of the most successful data leaders. In an April episode, Cindi was joined by Valerie Loga, CEO and founder of the DataLodge, and Jason Beyer, VP of Data and Analytics at Bridgestone Americas. The trio discussed one of the most important topics in the world of data: literacy. Specifically, both data and AI literacy.
The conversation is a must-listen, in my opinion, for anyone trying to build a stronger data-driven culture within their organization. Valerie opens the conversation by talking about the three key pillars of data literacy: mindset, language, and skills. These are the three things that will make people more confident and comfortable with working with data and AI, or working with any concept really. She also points out that with the skills pillar, the most important piece is critical thinking, and that as a society we are not naturally critical thinkers.
Jason discusses how he approached data literacy within his organization by offering a federated model of data where rather than compete with BI and analyst teams, they promote self-service amongst the masses and growth of those teams. What was interesting was the first department where things really took hold was HR. And because HR is connected throughout the entire organization, it had a high impact on getting traction across the rest of the organization.
As the conversation comes to a close, the group talks about how they are evolving data literacy to AI literacy. An interesting way they look at it is that it is not just a work skill but it is also a life skill, so it is really a gift to employees. They mention how similar steps taken to adopt data literacy will be needed to adopt AI literacy. Overall, I think this is an episode worth checking out if you are struggling to adopt or scale some form of literacy within your culture.
Last quarter, I helped lead a bakeoff evaluation between two of the most popular cloud data warehouse systems on the market: Snowflake and Azure Synapse. This bakeoff came out of numerous requests from customers and colleagues who were interested in learning more about how these particular systems compared to one another. Since SME is partnered with both Snowflake and Microsoft, we underwent the evaluation in an attempt to provide an unbiased comparison of these systems in the year 2024.
The bakeoff covers a wide range of topics from a technical and business perspective. This includes architecture, compute structure, development, extensibility, security, and costs. The goal of this bakeoff is to provide baseline information to help companies identify if either of these systems are a good or better fit for them. The link to the blog series is below. If you have deeper or more technical questions upon finishing the series, don’t hesitate to reach out to me to learn more!
At the end of March (I know, technically first quarter, just roll with it), the European Union gave the final approval for a groundbreaking law, the AI Act. This law would be the world’s first major law for regulating and installing comprehensive rules surrounding artificial intelligence. At its core, the law wants to highlight the importance of trust, transparency, and accountability in AI technology while allowing them to continue to innovate. The act will apply different scoring and regulation based on the risk or threat the tech poses to society. For instance, self-driving vehicles or medical devices will carry a high risk.
The act requires some time before its effects will actually be seen, as restrictions on general-purpose systems will not begin until one year after the act is in place, and currently available commercial systems like ChapGPT get a 36-month transition period. However, this is just the beginning in terms of seeing regulations and restrictions in the world of AI. Which, in my opinion, is expected. The rate of advancement of generative AI and LLMs over the past year has been staggering, to the point where AI and ChatGPT became household names. I don’t think that we are on the verge of anything out of iRobot or Terminator, but I do think that as we make these giant leaps of progress, we need to keep it in check. And at the end of the day, the EU AI Act is a fine. I do believe that as the AI continues to advance, there has to be stricter regulations and more accountability beyond just monetary value.
Anybody that knows me knows how passionate I am about music. I love listening to and discovering music of all genres, attending live music concerts and events, and learning more about how music is produced and distributed to the masses. To me, Spotify is one of, if not the greatest app to be released of all time. Having access to listen to whatever song I want whenever I am in the mood for it is amazing. But what about being able to create a song based on whatever mood I am in? This is where Suno comes in.
Suno is an AI application that allows anyone to create music by just feeding it a prompt. Similar to ChatGPT, Suno users enter in a text prompt of the song they want to hear, and the AI will generate two catchy tunes coupled with fitting lyrics in about 30 seconds. I was skeptical about it at first (as I typically am), but after trying a couple prompts with different genres, I am stunned at the output. The lyric generator comes across as cheesy, but the instrumentation and production it generates with these simple prompts is fascinating. What really blows me away is the singing. When prompted to give me metal and hardcore songs, the AI is able to generate screaming and emotional vocal infections that I have come to expect from these genres. If you put these songs in some Spotify genre playlists, you honestly wouldn’t be able to tell that something was different.
And while this is amazing, this is also problematic. Anyone with 200 characters can create music at their fingertips and pass it off as their own music. I have started to even see places like Reddit start to ban posting of AI-generated music in certain subreddits. It takes away from the talent and creativity of artists who spend countless hours and days working on their craft. Now do I think Suno is going to replace musicians? No, I think it is a cool concept and can have its place in creating things like corporate or commercial jingles. But it once again shows the power of AI and creates an issue where we will have to find ways to utilize it to promote creativity instead of outright replacing it.
Author’s Edit: Following writing this, it was announced that Suno is being sued by major record labels for training their AI on copyrighted music….a story we have heard before but in a different media medium.
While the first half may be over, 2024 has been exciting and innovative in the world of tech. These topics and events were just a few of the many highlights that we experienced so far. I am optimistic that the rest of the year will continue to bring us new technological innovations, exciting breakthroughs, and new ways for us to increase productivity. I am always keeping my eyes and ears open to learn new things, so if you want to discuss these topics or maybe another topic that you find interesting, you can message me on LinkedIn or email me at gbarrett@smesgroup.com. I hope you had a great first half of 2024 and have an even better second half!
Get In Touch With Us Today