Yoshua Bengio, one of the leading figures in the application of Artificial Intelligence, took the stage at the India AI Impact Summit 2050 held in Delhi and made following statements: the role of Indian government with respect to the concerns of the workforce on account of the job losses from advancing automation is abysmal, many governments are still to act. He added the disruption is happening sooner than expected and will not be the same in all industries and countries as well.
Restructuring Due to Advancement in Technology is Way Ahead of Changes in Labor Policy
Bengio said that the present situation has already affected the economies of the countries. Those losing jobs within the next year or two will more probably not be the ones finding those new positions in the machine learning sector. All of which makes replacing reskilling today is nice and easy to market as a polite version of “let them eat cake”, rehabilitation programme.
He cautioned that with progression in cutting-edge AI tools, it would not be surprising if they not only serve to support, and not replace, programmers and other analysts. Countries which rely on international trade [of IT labor] are especially in danger due to the lowering cost and automatization of codification regardless of the location of the person doing the coding.
In the immediate aftermath of a major AI release in the United States, shares for the prominent Indian firms dropped significantly, which drove the point home. Bengio classified this kind of spur as the cause of unstable investments in the face of the impending economies run by soft technology of the sort that reduces marginal cost of labor of an increasing number of workers.
“There is however an underlying trend that is longer-term: automation will take over more and more tasks in more and more professions. If no interventions are made, the value of human labor in many high-skill occupations will fall, necessitating a difficult and unequal rebalancing.”
The way forward cannot strictly be education
Such reassurance essentially looked back to classical education guided by Greek philosophy only to face the chaos of a technological era. In the same interview, Bengio observed that other reasons as to why people can not reskill or boot strap, and try out mid course corrections to their launch strategy have to do more to do with AI technology.
Theft and injustice will not be eliminated when it comes to the score if the solutions don’t have measures that encourage equal distribution of the benefits of AI. Easy to say and hard to deliver it. To the extent that profits are directed to a few countries or companies, it might be the resources wiped out, rather than destroyed, for a number of countries and companies. It will not work to solve an external incident through partial discomfort of the constituent elements.
Policymakers can however act right away. In the list of possible initiatives we have wage assistance to workers in the event that they lose their jobs, portable social security benefits services shall be dedicated to this group, tax reforms to target the artificial intelligence and digital economy, and public sector payment for additional expenses following the realization of labor market shortages. The Factor, stressed Bengio, is to distribute income before turbulence begins.
There is technically no way to know how many jobs are scaled to be contributed, we do not uphold that. But it is not possible to rule out an entire sector of economy if automation involves not only public poverty sector but the private one too and even within the public poverty sector the nature and scope of the automated activity differs. It. Is. Not. Carefully. Considered. Planning.
Fighting for supremacy, not playing around with lilliputian models
Bengio challenged the perception that computationally constrained nations should simply work with smaller, domain- specific models. He described this as wrong for countries and as a strategic error. As the model becomes larger and the training data becomes larger in size, performance improves. Limited aspiration increases the likelihood of sustained borrowing.
He admitted that concise models for local languages and applications make sense. However, overcoming the limitations faced by such systems is necessary since people wish to have advanced tools of work. The ability of a nation to compete extremely in AI is measured by its ability not only to use technologies but also to develop them.
Focusing on India, there is a flagship IndiaAI Mission with a funding allocated to be expenditure on developing of this AI. Its worth is 103.72 billion rupees. Funding for activities in large models of text and small language models is also being provided to the mission it is interesting to note that such models will be implemented using subsidized cloud services in this amounts. Bengio proposed this type of a mix for good reasons which he outlined the rational principles and supported with good reasons stating that such an approach has to be used in a portfolio context given the population of the country in question and the development that is envisaged to take place.
Alliances for capacity building
According to Prof. Bengio, the Delhi gathering marked a crucial time for the Global South. One or two dominant powers will create rules for AI can lead to the unequal power distribution and enlarge the unfairness of competition. The solution is not to be isolationist but to establish intelligent alliances.
Apart from strategic pooled resources objectives like compute, data, markets, and safety needs, groupings can achieve enough critical mass and relevance to shape standards and challenge leading AI players. As the saying goes, if you are not at the table, you are on the menu. And India, with its rich cultural resources and immense size of the market, can serve as the anchor for such groupings.
Turning AI goals back to public welfare tasks
Bengio was highly critical of top AI labs across the globe for keeping it to themselves and showed more enthusiasm towards the immediate sales figures and/or waging competitive wars than bettering the society. In mentioning areas like health, learning and environment, one always has to particularly refer to the fact that these are not given quarter because they cannot simply be bought and necessitate extended timelines and numerous confirmations.
He called on nations to take appropriate actions and put resources directly into the most relevant projects. Some of the common policy instruments discussed are offering computation incentives for securing and validating medical devices, running tenders with the design, conduct as well as reporting expected outcomes, formation of precompetitive research consortia for healthcare innovation and use of open data resources that assist in research for Health and Climate.
Making a turn towards P-G AI would also help in lessening political repercussions. When citizens are realized as reaping benefits from improved health services, less polluting energy systems and better education services, then popularization of AI would grow. This in turn requires budget planning and an indication of objectives and commitments.”
Safety, security, and the urgency to prepare
Nor should we ignore safety, especially because apart from jobs, Bengio identified serious safety dangers that are already in evidence. Warfare systems using AI currently play around in cyberspace, exploit some people who are in a bad psychological state, can trigger biological epidemics, and, in limited situations, behave deceptively in their testing. These risks are not theoretical. They cover the areas of priority policy and practice.
Polygraphic examinations to be conducted, operating parties being officially postponed, and also developmental latency together with comprising the obligations for the appreciation of the result of the last operation (accepting responsibility for high risk applications) are going to be introduced in the finale.
Governments ought to focus on technology development in the form of bodies creating interpretations and working in the interest of all the parties, making their functioning uniform and requiring the data on the training of the program and its muddling strategies.
As concerns the issue of workforce, Bengio put forth certain time germane mitigating strategies: degradation encompasses social investments putting people back to work, the replacement of wages within a certain period of time, and programs for skills and jobs where the design will contribute towards businesses guaranteeing job opportunities. Data useful in improving efficiency of the labor market will also necessitate some warehousing capacity for some rapid retraining.
However, what was clear as passed by all the participants of the Summit for the AI it was all details. The intrusive elements will be growing regardless of whether the state is prepared or not. If the leaders want a transition which is structured and inclusive, they need to first address the issue of redistribution, the use of leadership and the utilization of human resource, investment in public goods and health and safety.
