Our journey through 2023 and predictions for AI in 2024

From the august halls of power in Washington D.C., Brussels, and Beijing to kitchen tables around the world, AI has been the topic du jour in 2023. The discussion has been dominated by two opposing strains of thought. On one side, several influential technologists have pointed at the potential existential threats posed by superhuman AI. On the other side, techno-optimists have painted a picture of a technology that solves climate change and persistent diseases, seemingly overlooking any potential risks or downsides. Both groups agree that AI is a powerful technology but with highly divergent implications for society.

At Silo AI, our philosophy has consistently been grounded in a realistic and practical perspective
towards the technology, with a strong emphasis on placing humans at the core of AI. Moving forward, the evolution of AI is fundamentally dependent on the symbiotic relationship and collaboration between machine intelligence and human insight. As a true force multiplier akin to other general-purpose technologies, AI will impact the world in countless unforeseen ways. Since the founding of Silo AI in 2017, our journey through academic research and extensive client collaborations has revealed that adopting AI is a complex and nuanced process. Importantly, the real value of AI emerges in its embedded use – it elevates digital products and services, enhancing their capability and impact. Simply put, the value of AI is manifested in its ability to empower people, businesses, and societies through innovative, AI-integrated applications.

This approach continued to serve us well in 2023. To name a few highlights, we accomplished milestones on many fronts:

• launched the generative AI product arm SiloGen and the first models of the European multilingual open language model family Poro, with the intent to foster European digital sovereignty
• expanded to North America with the opening of an office in Vancouver and a partnership with Yoshua Bengio and 1,000 other academic AI researchers at Mila - Quebec AI Institute
• expanding in the Nordics through a strategic partnership with Combient and by
  acquiring their AI subsidiary Combient Mix, and expanding to the Benelux region by joining forces
  with Amsterdam-based Machine2Learn,
• expanded our offering in simulation and synthetic data through a strategic partnership with
  AILiveSim
• With 100 new Silonauts, our team now consists of more than 300 AI experts, including 150+ PhDs. In 2023, we also grew our revenue by ~50%, with a positive outlook on 2024.

With all of these developments, we strive to contribute to a world with more trusted and secure approaches to independent AI, in particular base models that enable tailoring for domain-specific needs. The development of base models aligned with European values is imperative for our digital sovereignty. We believe it’s essential that underlying models are based on data and information representing the citizens and organizations of the region, and overall compliance with regulation, data privacy and other vital concerns. And eventually we need sovereignty on how downstream applications and value creation happen. This way we can ensure digital sovereignty, while advancing technological development.

The milestones we have reached could not have been possible without the hard work, curious mindset, and willingness to do things differently exemplified by all Silonauts, and all our customers–most of whom have worked with us for several years. I am very grateful to all of you.

Looking ahead – AI in 2024

With the end of 2023 fast approaching, we believe that 2024 will be at least as exciting as 2023. The financial and intellectual investments in the development of the technology made over the last decades are paying dividends. The technology is becoming more mature, resource efficient, and simply more usable. There is a high number of motivated AI scientists, researchers, engineers, and hobbyists working on different parts of the value chain in universities, large tech companies, and smaller startups.

This perfect storm offers an opportunity to glean into the future. What will happen with generative AI in 2024? 

My first prediction is that in 2024, there will be a broad realization that as is usually the case, reality is more nuanced. More and more users and organizations will experience that the financial and societal value of AI technology will materialize through applications and products, not the technology itself. For generative AI, it means that the usage and value creation of Large Language Models (LLMs) will happen through models embedded into software products. LLM adoption will happen by a wide range of existing vertical and horizontal software vendors. AGI will not be achieved, but instead LLMs will create new features into existing software products, leading to increased value creation for both customers and vendors.

This leads me to my second prediction. The often repeated worry about job displacement is a relevant discussion in the longer term, but the technology is not quite there yet. Focus will shift to the jobs that AI creates. In other words, you are not likely to lose your job to AI, but you might lose it to someone who is adept at using AI.

My third prediction is that in 2024 an increasing number of organizations will realize that smaller, more affordable, more aligned and more specialized models make the most sense for a clear majority of LLM use cases. Priorities will include use-case specific performance, trustworthiness and cost. These will be empowered by open source models, bringing much needed transparency and trust to the field of AI.

While I do not believe today’s AI poses an existential risk, we should be mindful of the risks that do exist. We already have AI that passes the Turing test, and like with any tool AI can be used for good or bad. This is one of the key catalysts to policy makers on both sides of the Atlantic creating the first legal frameworks for AI. Compliance with these new rules will have to be a priority for companies throughout the value chain. In addition to ensuring one’s organization follows upcoming legal frameworks such as the EU’s AI Act, I believe organizations will invest more in efforts to address biases, enhance transparency and assess AI model's longevity, and to ensure LLMs are fit for their respective purposes. This is why my fourth prediction is that in 2024 we will see a clear increased focus on responsible AI practices. 

As we navigate the rapidly evolving landscape of AI, Silo AI’s focus will be on harnessing its value responsibly and sustainably, but with a sense of urgency that meets the potential of the technology. We are very excited to continue exploring the wide range of opportunities that AI brings us, and cannot wait to play our part in ensuring they become a reality!

All the best for the upcoming new year!

Peter