Twitxy Saved

Pradeep Goel

Verified account This account is verified. Learn more

@_PradeepGoel

Founder & CEO 30+ years experience in public health, finance, technology & insurance as CEO, COO, CIO and CTO & warrior for user rights.
pai3.ai Joined October 2017
738 Following 5K Followers 4.5K Posts

Latest posts 5

Pradeep Goel Verified account @_PradeepGoel
China's growing use of American AI models to train domestic systems shows how difficult it is to contain the spread of AI capabilities through chip restrictions alone.

A Reuters investigation found that Chinese researchers and companies are using outputs from models such as OpenAI's GPT-4 and o1 to develop training datasets for smaller domestic models. This practice, known as knowledge distillation, allows a less capable model to learn from the answers, reasoning examples and task demonstrations produced by a more advanced one.
3 replies 42 reposts 43 likes 599 views View on X
Pradeep Goel Verified account @_PradeepGoel
The introduction of AI into air traffic management creates a central dilemma: how to capture the speed and data processing benefits of automation without eroding the human oversight required to prevent catastrophic mistakes.

The FAA's new SMART system is an interesting case. Developed with Air Space Intelligence, it brings together 200 data streams, including weather forecasts, flight paths, traffic flow and controller staffing information, to help specialists anticipate congestion and identify better routes before problems arise.
5 replies 51 reposts 56 likes 781 views View on X
Pradeep Goel Verified account @_PradeepGoel
Drug discovery and clinical development operate on very different timelines. Finding a promising molecule or identifying a potential treatment target is an important scientific achievement, even though it does not always establish that the treatment is safe and effective in people.

Before a potential treatment can reach patients, it must undergo preclinical testing, clinical trials and regulatory assessment. Each stage takes the time researchers need to establish that the treatment is safe, demonstrate that it works in humans and carefully evaluate its risks. These requirements cannot be bypassed by making computational research faster.

Computational tools could help researchers identify targets, analyze biological data, design molecules and improve parts of the development process, making important contributions to medicine.

However, we should be careful about turning faster discovery into promises of rapid cures. The measure of progress is if these advances produce treatments that are safe, effective and accessible to the people who need them.
3 replies 51 reposts 54 likes 669 views View on X
Pradeep Goel Verified account @_PradeepGoel
Many forms of intellectual work depend on individual expertise, practice and time. Research, analysis and design are seen as crafts because producing good work required considerable human effort.

This economic model is now being challenged and the consequences will reach past productivity. Businesses are already rethinking staffing and professional services are facing new competition. People whose livelihoods depend on these activities will have to navigate a difficult transition.

We should be honest about the scale of this change while making sure that the gains do not accrue only to the owners of the technology. The people whose work is being transformed deserve a place in the conversation about how this new economy is organized.
1 replies 52 reposts 56 likes 677 views View on X
Pradeep Goel Verified account @_PradeepGoel
Human beings have always explored the world through the limits of our senses, language and ability to reason. Even the questions we ask are shaped by what we already know how to observe and describe.

There may be patterns in nature, mathematics and complex systems that are difficult for the human mind to recognize because they do not fit familiar ways of thinking.

Advanced computational systems could help us investigate some of these unfamiliar relationships. They may identify structures in large datasets or connections that would take us much longer to discover.

Discovering a pattern is only part of the work. We need methods to test it, understand its significance and establish whether it reflects something real. I find this possibility exciting because it could expand the boundaries of scientific inquiry beyond what we have been able to explore so far.