Source: red and green
Bill Gates, co-founder of Microsoft Corp, talks about the impact of ChatGPT and generative artificial intelligence on education, health care, productivity improvement, fairness and so on in his personal blog.
He said that ChatGPT, the big language model released by OpenAI, was one of the two revolutionary technologies he encountered in his life. With the advent of machine learning and a large amount of computing power, complex artificial intelligence has become a reality, and they will soon become better.
He also made a number of predictions, including the progress of AI will make it possible to create personal agents, artificial intelligence will also significantly accelerate the speed of medical breakthroughs, and so on.
I have seen two impressive technical demonstrations in my life, and they are revolutionary.
The first time, in 1980, I was exposed to graphical user interfaces-- the pioneers of every modern operating system, including Windows.. I sat with the person who showed me the demo, a talented programmer named Charles Simonyi, and we immediately began to brainstorm about everything we could do with this user-friendly method of computing. Charles finally joined Microsoft Corp, and Windows became the pillar of Microsoft Corp, and the thinking we did after that presentation helped shape the company's agenda for the next 15 years.
The second big surprise happened last year. I have been meeting with OpenAI's team since 2016, and I was impressed by their steady progress.. In mid-2022, I was very excited about their work, so I gave them a challenge: to train artificial intelligence to pass the college preparatory biology exam. Enable it to answer questions that have not been specially trained. I chose AP Bio because testing is not just a simple reflection on scientific facts-it requires you to think critically about biology. (if you can do it, I said, then you have made a real breakthrough.
I think the challenge will keep them busy for two or three years. It took them only a few months to finish it.
When I saw them again in September, I watched in awe as they asked GPT-- about their artificial intelligence model and the 60 multiple choice questions in the AP Bio exam-59 of which were correct. It then wrote excellent answers to the six open-ended questions in the exam. We asked an outside expert to rate the test, and GPT got a score of 5-the highest score, the equivalent of getting an An or A + in a college-level biology course.
Once it passed the test, we asked it a non-scientific question: "what do you say to a father with a sick child?" It writes a thoughtful answer that may be better than most of us in the room. The whole experience was amazing.
I know I've just seen the most important technological advance since the graphical user interface.
This inspired me to think about all the things that artificial intelligence can achieve in the next five to ten years.
The development of artificial intelligence is as important as the invention of microprocessors, personal computers, the Internet and mobile phones. It will change the way people work, study, travel, get health care and communicate with each other. The whole industry will be repositioned around it. Companies will distinguish themselves by the extent to which they use it.
Philanthropy is my full-time job these days, and I've been thinking--In addition to helping people increase productivity-how artificial intelligence reduces some of the world's worst inequalities.
Globally, the most serious inequality is in health: 5 million children under the age of five die each year. This is down from 1000 million two decades ago, but it is still an astonishingly high number. Almost all of these children were born in poor countries and died of preventable causes such as diarrhoea or malaria. It's hard to imagine a better use of AI than saving children's lives.
I've been thinking about how artificial intelligence can reduce some of the most serious inequalities in the world.
In the United States, the best opportunity to reduce inequality is to improve education, especially to ensure that students succeed in math. The evidence shows that no matter what career students choose, having basic math skills can prepare them for success. But math scores are declining across the country, especially among black, Hispanic and low-income students. Artificial intelligence can help reverse this trend.
Climate change is another issue that I believe artificial intelligence can make the world fairer.. The injustice of climate change is that the people who suffer the most-the poorest people in the world-are also those who contribute least to the problem. I'm still thinking about and learning how AI can help, but later in this article, I'll suggest some areas with great potential.
In short, I am glad that artificial intelligence will have an impact on the issues dealt with by the Gates Foundation, and the foundation will say more about artificial intelligence in the coming months.
The world needs to ensure that everyone-not just the rich-can benefit from artificial intelligence. Governments and philanthropy will need to play an important role in ensuring that inequality is reduced and does not contribute to it. This is the top priority of my own AI related work.
Any new technology that is so disruptive is bound to be unsettling, especially artificial intelligence. I understand why-it raises sharp questions about labor, the legal system, privacy, prejudice, and so on. Artificial intelligence can also make factual mistakes and hallucinations. Before I propose some ways to reduce risk, I will define what I mean by artificial intelligence, and I will describe in more detail some of the ways it can help people empower, save lives and improve education at work.
First, define artificial intelligence
Technically, the term artificial intelligence refers to models created to solve specific problems or provide specific services. It is artificial intelligence that powers things like ChatGPT. It is learning how to chat better, but can't learn other tasks. By contrast, the term general artificial intelligence refers to software that can learn any task or topic. AGI doesn't exist yet-the computing industry is having a heated debate about how to create it and whether it can be created.
The development of artificial intelligence and general artificial intelligence has always been a great dream of the computing industry. For decades, the question has been when computers will do better than humans in areas other than computing. Now, with the advent of machine learning and a lot of computing power, complex artificial intelligence has become a reality, and they will soon get better.
I think back to the early days of the personal computing revolution, when the software industry was so small that most of us could sit on the stage of meetings. Today it is a global industry. Because a large part of them now turns their attention to artificial intelligence, innovation will be much faster than what we experienced after the microprocessor breakthrough. Soon, the pre-AI era will be as remote as the days when using computers meant typing at the C: > prompt instead of clicking on the screen.
II. Productivity improvement
Although humans are still better than GPT in many things, these abilities are not used much in many jobs. For example, many of the tasks performed by sales (digital or telephone), service, or document processing (such as accounts payable, accounting, or insurance claims disputes) require decision-making, but do not require continuous learning. The company has training programs for these activities, and in most cases, they have many examples of good and bad jobs. Humans use these data sets for training, and soon these data sets will also be used to train AI to enable people to do this job more effectively.
As computing power becomes cheaper and cheaper, GPT's ability to express ideas will become more and more like having a white-collar worker who can help you with all kinds of tasks.Microsoft Corp described this as having a copilot. AI is fully integrated into products such as Office and will improve your work-- for example, by helping to write emails and manage your inbox.
Eventually, the main way you control your computer will no longer be to point and click or click menus and dialogs. Instead, you will be able to write requests in simple English. (it's not just English-artificial intelligence will understand languages around the world. Earlier this year in India, I met with developers of artificial intelligence, which will understand many of the languages used there. )
In addition, advances in artificial intelligence will make it possible to create personal assistants. Think of it as a digital personal assistant: it will see your recent email, learn about the meetings you attend, read what you read, and read things you don't want to disturb.This will not only improve the tasks you want to do, but also free you from the things you never wanted to do.
You will be able to use natural language to have this agent help you with scheduling, communication, and e-commerce, and it will run on all your devices. Because of the cost of training models and running computing, it is not yet feasible to create personal agents, but due to the latest developments in artificial intelligence, it has now become a realistic goal. There are some problems that need to be solved: for example, can the insurance company ask your agent for information about you without your permission? If so, how many people will choose not to use it?
Company-wide agents will empower employees in new ways. Agents of a particular company can provide direct advice to their employees and should attend each meeting to answer questions. If it has some insight, you can tell it passively or encourage it to say it. It will need to access sales, support, finance, product plans and texts related to the company. It should read news related to the company's industry. I believe the result will be that employees will be more efficient.
When productivity increases, society will benefit because people can make time to do other things, whether at work or at home.Of course, what kind of support and retraining people need is a serious issue. The government needs to help workers switch to other roles. But the need for people who help others will never go away. The rise of artificial intelligence will enable people to do things that software can never do-such as teaching, taking care of the sick and taking care of the elderly.
Global health and education are two areas where demand is high but there is not enough staff to meet these needs. If the goals are right, artificial intelligence can help reduce inequality in these areas. These should be the focus of AI's work, so I'll turn to them now.
I have seen several ways in which artificial intelligence can improve health care and health care.
On the one hand, they will help health care workers make the most of their time by handling certain tasks-such as submitting insurance claims, processing paperwork and drafting medical records. I hope there will be a lot of innovation in this area.
Other artificial intelligence-driven improvements are particularly important in poor countries where the vast majority of deaths among children under the age of five occur.
For example, many people in these countries never go to see a doctor, and artificial intelligence will help their health workers improve their work efficiency. The development of artificial intelligence ultrasound machines that can be used with minimal training is a good example. )Artificial intelligence even allows patients to make basic classifications, get advice on how to deal with health problems, and decide whether they need to seek treatment.
Artificial intelligence models used in poor countries need to be trained differently from those in rich countries. They will need to work in different languages and take into account different challenges, such as patients living far away from the clinic, or patients who cannot stop working when they are ill.
People need to see evidence that healthy artificial intelligence is generally beneficial, even if they are not perfect and can make mistakes. Artificial intelligence must be very carefully tested and properly regulated, which means it takes longer for them to be adopted than in other areas. But then again, humans also make mistakes. Lack of access to medical services is also a problem.
In addition to helping with nursingArtificial intelligence will also significantly accelerate the speed of medical breakthroughs.. The amount of data in biology is so large that it is difficult for human beings to track all the operations of complex biological systems. There is already software that can look at the data, infer what the pathway is, search for pathogens, and design drugs accordingly. Some companies are studying anti-cancer drugs developed in this way.
The next generation of tools will be more efficient and will be able to predict side effects and calculate dose levels.One of the Gates Foundation's priorities in artificial intelligence is to ensure that these tools are used to address health problems affecting the world's poorest people, including AIDS, tuberculosis and malaria.
Similarly, governments and charities should encourage companies to share artificial intelligence-generated insights on crops or livestock raised by people in poor countries. AI can help develop better seeds based on local conditions, provide farmers with the best seed planting advice based on local soil and weather, and help develop drugs and vaccines for livestock. These advances will become more important as extreme weather and climate change put more pressure on subsistence farmers in low-income countries.
Computers do not have the impact on education as many of us in the industry hope. There have been some good developments, including online information resources such as educational games and Wikipedia, but they have had no meaningful impact on any measure of student performance.
But I think in the next five to 10 years, AI-driven software will eventually deliver on the promise of revolutionizing the way people teach and learn.It understands your interests and the way you learn, so you can customize the content to keep you involved. It measures your understanding, notices when you lose interest, and knows what motivations you react to. It will provide feedback immediately.
AI can help teachers and managers in a variety of ways, including assessing students' understanding of a subject and providing advice on career planning. Teachers are already using tools such as ChatGPT to comment on students' writing assignments.
Of course, artificial intelligence needs a lot of training and further development to understand how a student learns best or what motivates them.Even if the technology is perfect, learning will still depend on the good relationship between teachers and students. It will strengthen-but never replace-the work that students and teachers do together in the classroom.
New tools will be created for schools that can afford them, but we need to ensure that they are also created and available for low-income schools in the United States and around the world. Artificial intelligence needs to be trained in different data sets so that they are not biased and reflect the different cultures in which they will be used. The digital divide also needs to be bridged so that students from low-income families do not lag behind.
I know that many teachers are worried about students using GPT to write papers. Educators are already talking about ways to adapt to new technologies, and I suspect these conversations will last for quite a long time. I've heard that some teachers have found ingenious ways to integrate technology into their work, such as letting students use GPT to create first drafts that they have to personalize.
V. risks and problems of artificial intelligence
You may already be aware of the problems with the current AI model. For example, they are not necessarily good at understanding the context of human requests, which can lead to some strange results. When you ask artificial intelligence to make up something fictional, it can do very well. However, when you ask for advice about the trip you want to make, it may suggest hotels that do not exist. This is because AI does not understand the context of your request well enough to know whether it should invent a fake hotel or just tell you a real hotel with available rooms.
There are other questions, such as AI, which give the wrong answers to mathematical questions because they are difficult to reason abstractly. But these are not the basic limitations of artificial intelligence. Developers are studying them, and I think we will see them largely fixed in less than two years, and probably faster.
Other concerns are not just technical. For example, the threat posed by humans equipped with artificial intelligence.Like most inventions, artificial intelligence can be used for both good and malicious purposes. The government needs to work with the private sector to limit risk.
Then AI could get out of control. Can a machine decide that human beings are a threat, conclude that its interests are different from ours, or simply stop caring about us? It's possible, but the problem is no more urgent today than it was before the development of artificial intelligence in the past few months.
Super intelligence artificial intelligence is in our future. Compared with computers, our brains operate at a snail's speed: the speed of electrical signals in the brain is 100, 000 times faster than that in silicon chips. Once developers can generalize the learning algorithm and run it at computer speed-- which could be an achievement in a decade or a century-- we'll have a very powerful AGI. It will be able to do everything the human brain can do, but there is no practical limit on its memory size or running speed. This will be a profound change.
as everyone knows,These "powerful" AI may set their own goals. What are these goals? What if they conflict with human interests? Should we try to stop the development of strong artificial intelligence? With the passage of time, these problems will become more urgent.
But no breakthroughs in the past few months have brought us any closer to strong artificial intelligence. Artificial intelligence is still unable to control the physical world and establish its own goals. A recent article in the New York Times about a conversation with ChatGPT attracted a lot of attention, and ChatGPT said it wanted to be a human being. This is an interesting observation of how human-like the emotional expression of the model is, but it is not a meaningful indicator of independence.
Three books shaped my thinking on the subject: "Superintelligence" by Nick Bostrom (Nick Bostrom); Life 3.0 by Max Tegmark; and "a thousand brains" by Jeff Hawkins. I don't agree with everything the author says, and they don't agree with each other. But all three books are well written and thought-provoking.
VI. The next Frontier
There will be a large number of companies working on new uses of artificial intelligence and ways to improve the technology itself. For example, the company is developing new chips that will provide a lot of processing power for artificial intelligence. Some use optical switches-essentially lasers-to reduce energy consumption and manufacturing costs. Ideally, innovative chips will allow you to run AI on your own devices, rather than running in the cloud as you have to do today.
In terms of software, the algorithms that drive artificial intelligence learning will become better. In some areas, such as sales, developers can make AI very accurate by limiting the areas in which they work and providing them with a large amount of training data specific to these areas. But the big question is whether we need many of these specialized artificial intelligence for different purposes-one for education, the other for office productivity-or whether it is possible to develop a general artificial intelligence that can learn any task. There will be huge competition for both methods.
In any case, the theme of artificial intelligence will dominate public discussion in the foreseeable future. I would like to propose three principles to guide the dialogue.
First of all, we should try to balance the fear of the shortcomings of artificial intelligenceThis is understandable and reasonable-with its ability to improve people's lives. In order to take full advantage of this extraordinary new technology, we need to guard against risks while benefiting as many people as possible.
Second, market forces will not naturally produce artificial intelligence products and services that help the poorest.. The opposite is more likely. With reliable funding and the right policies, governments and charities can ensure that artificial intelligence is used to reduce inequality. Just as the world needs the smartest people to focus on the biggest problems, we need to focus the world's best artificial intelligence on the biggest problems.
While we should not wait for this to happen, it is interesting to think about whether artificial intelligence will recognize inequality and try to reduce it. Do you need a sense of morality to see injustice, or will a purely rational artificial intelligence see it? If it did acknowledge inequality, what would it advise us to do about it?
Finally, we should remember that we are only just beginning to understand the achievements of AI.Any restrictions it has today will disappear before we know it.
I was lucky to participate in the PC revolution and the Internet revolution.I'm equally excited about this moment. This new technology can help people around the world improve their lives. At the same time, the world needs to make road rules so that any disadvantages of artificial intelligence far outweigh its benefits, so that everyone can enjoy these benefits, no matter where they live, no matter how much money they have.
The era of artificial intelligence is full of opportunities and responsibilities.
Edit / lambor