Over the years, we have seen the rise of new types of AI, including generative and agents. How do you see 2025 and beyond developing AI?
Aye hai While words like “generic” and “agentic” are helpful in simplifying technology for the general public, they can also be misleading. These abilities, such as natural language generation (NLG), exist for some time and represents just one part of the very comprehensive AI toolbox.
As hardware continues to improve, AI will be more capable, more specific and more deeply embedded in our daily life, similarly the Internet gradually became a fundamental technique. A strong example is medical AI, which is emerging as a new standard of rapid care. While some are early adoptives, we are seeing a wave of sharp followers. While choosing doctors, hospitals and insurance providers, patients have started expecting AI-operated capabilities. Physician’s acceptance in 2019 has increased from about 35% to about 70%, which is an important cultural change.
Under the Trump administration, many market participants are expecting to change the rules around AI and have encouraged that the US adopts a slow view than European regulators. Do you think how will the rules change during this administration?
This administration appears to be practical and assistant to American trade. The overgrowth slows down risk innovation, especially with acute AI competition coming from China and Russia. I would expect the administration to return the US-based AI companies as a strategic assets like Delorian AI.
The regulatory approach of the European Union has stopped its own technical field, in many ways. Major American tech firms have faced an important regulator headwind in Europe, and the region’s AI industry has fought to remain competitive globally. It should work as a careful story for us.
Many concerns labeled in the form of “AI Ethics” have already been covered under the existing data privacy laws. Instead of creating new, overlapping rules, government agencies should focus on implementing when implemented in advance.
In the end, I would strongly recommend that the administration seek guidance from real physicians, who build and use AIs every day, rather than relying only on commentators or academics that can be removed from the real world applications.
When you think about the global AI rules, how can we ensure that the rules will not interrupt innovation and development?
From their nature, rules prevent innovation and development. However, I believe that the basic railings we need are already established through the current data law. Enforcement, not expansion, should be focused.
I will encourage the US government to support domestic AI companies in many major areas:
- Supply chain security: Ensure that we have the necessary materials for hardware – rare earth, chips, servers.
- IP Conservation: Protection of American innovation. If foreign actor IP is engaged in theft, his US-based representatives should be held accountable.
- R&D incentive: The current R&D tax credit has come down. We need a more meaningful innovation for AI innovation.
- Talent Strategy: In short -term, expand the H1B visa. In the long term, we should strengthen STEM education and ensure that our universities are producing A-Taiyar talent.
Finally, we should make it easier for government agencies and private companies to adopt AI equipment. In this way we remain competitive.
In myself Recently Tradtock DiscussionYou called hardware and maximum server capacity requirement as AI continues to develop. What are your expectations for storage growth in the next year and next decade?
The only honest answer is: yes, and development will be exponential.
From a national security and economic point of view, we should secure the raw materials and access to efficient labor required to manufacture and operate chip and server infrastructure. My colleagues and I am already searching for places that provide the power capability required to host these server forms.
This demand offers a compelling opportunity to integrate renewable energy sources in the AI infrastructure. For example, old factory sites in New England can be revived using hydroelectric power. The region has a tremendous capacity of permanent growth.
You have also mentioned that AI cannot be biased itself. Can you expand on how companies can ensure that there is no bias in the dataset used to make their AI models?
This is correct, AI, as a machine, is not naturally biased. Any bias comes from the data on which it is trained. And this is the place where things become complicated.
First of all, companies should regularly audit their model to ensure that there is no bias related to the legally protected classes. We already have rules that are needed. Second, well -trained scientists who follow scientific method understand the importance of designing balanced datasets from the beginning.
It is also important for customers to ask the correct questions of your vendors. Transparency is important in model development and training data.
He said, sometimes data reflects a naturally homogeneous population. For example, a model trained on data from Iceland, where the population is relatively identical, if applied to diverse areas such as Orlando, it cannot perform well. It is not a bias in the model, but a mismatch in the training data vs. application context.
What can companies and policy makers do today to prepare for the next wave of AI innovation?
Companies need to invest in AI literacy at the leadership level. Many times, we see COI handing CIOs handing over AI decisions that are experts in it, but not in AI. This is an important misleading. You need decision making who understand the unique nature of AI technology.
In addition, there is no need to strengthen the wheel – buy Siddha AI product. The construction of custom solutions at home may not be understood for some industries, such as healthcare, where AI is not the main qualification.
For policy makers, it is important to find input from primary sources, who are manufacturing and using AI, not only theorists or strategists. Real world physicians provide the most ground, actionable insight. And most importantly, policy makers should focus on enabling and nourishing the US AI industry, rather than to regulate it.
