The Intersection of Decreasing AI Costs and Notetaking in 2022

The Intersection of Decreasing AI Costs and Notetaking in 2022

In recent years, we have witnessed a significant decrease in the costs associated with artificial intelligence (AI) training and inference. This trend has been fueled by various factors, including algorithmic improvements and the decreasing prices of GPUs. One company that has embraced this trend is Databricks, which recently acquired Mosaic, a company that focuses on making AI training cost-effective for businesses.

Mosaic's vision aligns perfectly with Databricks' goal of helping companies rapidly adopt machine learning to stay ahead of the competition. The training costs for AI models have decreased by a staggering 10 times in less than a year. For example, it now costs $50k to train stable diffusion and $200k to train a high-quality LLM. MosaicML, in particular, has demonstrated its ability to train stable diffusion models for just $50k, as opposed to the $600k it cost Stability.

There are two primary drivers behind this trend. Firstly, companies like MosaicML have made significant algorithmic improvements, resulting in more efficient and cost-effective training processes. Secondly, the prices of GPUs have decreased by three times in just three years. For instance, the cost of a Nvidia T4 GPU for one hour was $0.95 in August 2019, but it is now only $0.35 per hour. These advancements in both algorithms and hardware costs have paved the way for more accessible and affordable AI training.

The decreasing costs of AI training have broader implications for the industry. As the cost barrier lowers, we can expect to see more model providers entering the market, leading to increased competition at the model layer. This, in turn, should put pricing pressure on closed-source model providers and potentially encourage more companies to opt for open-source alternatives.

While AI costs are decreasing, another area that has seen significant developments is notetaking. In 2022, notetaking has become an essential activity for many digital writers and knowledge workers. The act of notetaking is driven by the desire to observe and capture valuable insights and information.

For some, notetaking has become a daily practice, with writers like Arvid Kahl writing an essay every day. This consistent effort has transformed their approach to digital writing and enabled them to see themselves in a new light. Notetaking has become a catalyst for personal growth and self-reflection.

In the realm of notetaking apps, there has been an interesting shift in preferences. Scrintal, a bootstrapped founder podcast, has emerged as a popular choice among digital writers and knowledge workers. Many have replaced their combination of Roam Research and Evernote with Scrintal, considering it a pivotal moment in their notetaking journey.

Additionally, Glasp, a social highlighter and community platform, has gained attention for its unique concept. Users can share their highlights with other writers and knowledge workers, fostering collaboration and idea exchange. This platform's potential integration with Readwise, a popular notetaking app, is an exciting prospect for the future.

When it comes to notetaking, the key is to invest in apps that save time rather than consume it. The goal is not to complicate things but to simplify them. By prioritizing efficiency and simplicity, writers can focus on the core essence of notetaking: writing itself. Consistency and intentionality are vital in this practice, as emphasized by Andy Sporring, who has written 100 notes to date.

However, amidst these advancements in notetaking apps, Evernote has faced a decline in trust and confidence. Many users have expressed disappointment with the once highly regarded Swiss army knife of apps. While the exact reasons for this loss of trust may vary, it serves as a reminder that even the most established platforms can falter.

In conclusion, the decreasing costs of AI training and the evolution of notetaking in 2022 are two distinct yet interconnected trends. As AI training becomes more cost-effective, it opens doors for businesses to leverage machine learning and gain a competitive edge. Simultaneously, digital writers and knowledge workers are exploring new and improved notetaking apps that simplify the process and enhance collaboration.

To navigate these trends effectively, here are three actionable pieces of advice:

1. Stay updated on the advancements in AI training costs and algorithms. By staying informed, you can make informed decisions when choosing model providers and embrace more cost-effective options.

2. Explore different notetaking apps and find one that aligns with your workflow and saves you time. Prioritize simplicity and efficiency to enhance your writing and knowledge management practices.

3. Embrace consistency and intentionality in your notetaking journey. Set aside dedicated time each day to write and observe, allowing yourself to grow as a digital writer and knowledge worker.

By incorporating these suggestions into your approach, you can make the most of the decreasing costs of AI and optimize your notetaking practices in 2022 and beyond.