How will artificial intelligence develop in 2023?

Artificial Intelligence (AI) | From large organizations to startups to independent research labs, machine learning teams are working hard every day to build and deploy models that improve everyday life. As the year comes to an end, it’s good to stop and see their progress.

In many ways, the year 2022 begins and ends with generative AI. Last year saw the launch of Dall-E 2 OpenAI (April), but also Medjourney (April), from spreading stability 2 From Stability AI (Aug) and from chat (November). The technical breakthroughs that are made almost every week, whether transforming robots or improving the study of genomes, earn less headlines, but they are just as revolutionary.

For most machine learning professionals, the problem today is not a lack of research, new tools, or techniques, but rather the fact that they just can’t keep up with all of these advances!

Before looking ahead, here’s a review of the past year.

• AI stocks will decline before they get better: TRUE. In general, the impact of AI behaving in a discriminatory manner will probably not be fully known, but this year leaves a lot to be desired. For example, it is now clear that harmful biases lurk in large-scale text-to-image and language models. Model bias also continues to emerge in more mainstream models, affecting everything from health outcomes to compensation claims. Finally, there is still a long way to go to ensure better diversity in AI recruitment and ethics.
• Companies will stop deploying AI blindly: partly true. With the adoption of machine learning monitoring accelerating and market leaders emerging, the reality is that many teams still lack a monitoring plan to early detect and diagnose problems with models in production. This is especially true for teams that have deployed computer vision and natural language processing models, as the tools for monitoring things like integration drifts are still very new.
• Peak Scientist Citizen: Partially True. While the adoption of low code tools is still a factor in the democratization of data science, it has been overshadowed by the revolution that is emerging around large language paradigms with text as the universal interface.
• The machine learning infrastructure ecosystem will become more dense and complex: right. With increased investment in machine learning and AI infrastructure tools this year, the ecosystem is getting more crowded. Overall, 85.7% of big data experts and machine learning engineers say they still sometimes “struggle navigating a confusing and disorganized machine learning infrastructure ecosystem.”
• The number of machine learning engineering positions will exceed the number of available talent, resulting in a talent shortage: right. While recent layoffs are a concern, the past year was particularly marked by labor shortages across the economy, particularly in data science and machine learning.

In 2023, here are the AI trends to watch.

1. Generative AI will become mainstream (and so will its growing pains)

Generative AI captures audience imaginations like some of the technical advances that have taken place since the advent of cinema more than 100 years ago. With powerful apps like githubAnd the copilot where chat, which has already proven itself, many companies are eager to adopt this technology on a larger scale. However, generative AI is still uncharted territory. There will be many elements to be clarified in 2023, particularly with regard to prejudice, copyright, scalability, security and how to monitor this new technology. In short, generative AI will require a network, and one will have to build that network.

2. Economic uncertainty will be a major driver of the machine learning infrastructure market

Artificial intelligence is likely to gain more importance as inflation and economic turmoil put pressure on companies to become more efficient and productive. Due to changing priorities, this is the end of the days when centralized machine learning teams took months or even years to build and maintain feature stores or internal monitoring tools. Buy rather than build is likely to become more popular, not least because teams need to prioritize projects that increase company revenue in the near term. Given the economic environment, purchasing pressures, or even layoffs, are not likely to affect machine learning teams in some industries. In this context, only tools MLOps The strongest groups that bring real added value to teams will flourish. Attendez-vous à ce que des outils tels que les plateformes d’orchestration, qui mettent en lumière certaines hypothèses dépassées concernant la connexion de nombreux outils d’apprentissage automatique disparates, ainsi que les leaders hors categorie aient du mal à lever des fonds ou à go bankrupt

3. The most successful platforms will attack traditional players

That’s what happened with DevOps And that’s what’s happening now with MLOps: in technical areas, the best platforms tend to win. Given the complexity of modern machine learning, machine learning teams need deeper tools at every stage of the model lifecycle. As a result, platforms offering so-called integrated solutions, which emerged a decade ago to support the work of big data experts and machine learning teams, are losing developer share and facing layoffs. Even the big players like Amazon (with sigmacker) and Google (with Vertex) does not currently reflect the technical depth required for each part of the machine learning lifecycle, although wave consolidation could change that.

4. Working with unstructured data will not be optional

Unstructured data is everywhere. According to several estimates, 80% of the data generated worldwide is in the form of unstructured images, text, video or audio files. In recent years, some of the most powerful modern applications in terms of machine learning are exploiting unstructured data. A machine learning platform that is not designed to handle unstructured use cases may not be relevant or have limited growth potential. At the same time, machine learning teams that find ways to take advantage of computer vision or natural language processing models, even if they are just applying a pre-trained model to a situation that is barred from commercial use, can find new competitive advantages.

In conclusionWhile not every forecast looks bright or positive, there are plenty of reasons to be optimistic about the future of AI and machine learning teams. It remains only to hope that everyone can start 2023 with the optimization of this sector in mind!

Translated article from Forbes US – Author: Aparna Dinakaran

<< لقراءة سورالسو: الذكاء الاصطناعي في خدمة استدامة الشركات >>>