Shulou(Shulou.com)11/24 Report--

Feb.13 news, just after the hype surrounding the meta-universe has passed, the emergence of OpenAI's AI chat robot ChatGPT and other generative AI (AIGC) has attracted the attention of many long and short bets on AI, as technology companies are chasing "the next big thing."

Meta-universes come with extensions of NFT, cryptocurrency and blockchain, but for now, all of these look like symptoms of the peak of the hype cycle. But one thing is certain: silicon valley needs the next big thing because the industry is undergoing a strategic contraction that is very different from the past decade, and technology-led economic layoffs and cost cuts are now becoming the norm in an industry accustomed to relying on investors to write blank checks.

Not long ago, CNBC's Technology Executive Committee virtual forum brought together tech executives from companies across multiple economic sectors, including many that use AI but haven't developed it themselves, such as retail, media, legal, agriculture, and logistics. At the conference, AI supporters and skeptics were equally numerous, and they were divided into several groups to discuss the trend of people exploding on ChatGPT and to distinguish hype from reality as much as possible.

A key point raised by several executives was that while the meta-universe remains a vague concept for many, AI is now accelerating certain technology processes that have been used for 20 years across a wide range of functions in companies, including software engineering, finance, operations, legal, logistics, and creative. Even skeptics of the latest AI hype recounted numerous examples during the conference of AI being embedded in more efficient business processes.

The market is clearly taking the latest advances in AI seriously, as evidenced by Microsoft and Google's fierce competition over AI in search engines this week. Google's attempt to respond to Microsoft's challenge with its AI chatbot Bard was seen as a "rushed and botched" effort that could only amplify the risks to its dominance in search. Since then, Google shares have fallen about 13% in two days.

For his part, Microsoft CEO Satya Nadella clearly enjoys seeing competitors in anxiety. In a recent interview, he said: "From now on, the gross margin of the search business will always decline. There's still a lot of margin in search, which is incremental for us, but not for Google, and they have to fight to defend it. "

Microsoft's AI efforts are making rapid progress, with Reid Hoffman, a board member, saying last year: "Miracles do happen. The AI Hoffman is discussing, called Copilot, is now taking over up to 40 percent of code-writing tasks from human software engineers, but Hoffman predicts: "AI will change all industries." So everyone has to think about it, not just in the field of data science. In everything we do, we will have amplification tools that will become a reality in the next 3 to 10 years and become the baseline for everything we do. "

The latest discussions suggest that AIGC and AI more broadly will kick off a new era of Internet search. A cloud executive said at the conference: "This is a very, very important change, and people even compare AI to changes in cloud computing, mobile and even the Internet. And I think AI will be 100 times bigger than all of that put together, comparable to the Industrial Revolution. I think AI has a lot in common with the Industrial Revolution. Companies that do not immediately embrace it are likely to end up being eliminated. "

For tech leaders willing to integrate AIGC into their operations and consumer-facing services and products, the following are the most likely benefits and risks:

Many advances in AI in recent decades have shown that computational power can be used to master a single complex task, a chess game, or a recommended algorithm for Netflix or TikTok. ChatGPT, on the other hand, redefines the limits of what machines can learn, and most executives seem to agree that another industrial revolution-like process is brewing.

Deep language learning models currently being developed and rolled out will eventually span all departments and take over the manual work of all current functional teams. One executive gave the example of stock analysis. "We can use the AIGC in the financial sector," he said. We will read 5000 balance sheets in seconds, be able to extract all financial information, calculate risk scores, and be able to make decisions about portfolio risk. "

"If you can train deep language-learning models, your solution could have a profound impact," says another executive. "While traditional AI can only solve problems in the 'deep analytics domain,' this new AI brings these capabilities into the creative economy." He added: "How do you see the creative economy? This has profound implications for both functional and creative work. "

Often, the common argument against AI is that it will be a job killer, but executives in all walks of life have been refuting this for years. This is not the case, they argue, and AI will simply replace repetitive or mundane tasks that humans do not want to do, allowing them to do more important work. While there's no definitive answer, that's the main message from the executives. For example, one executive said that manual laboratory operators used to have to work hard in front of microscopes and images, but these jobs have now been replaced.

Another executive who works with lawyers and accountants said the current view is that AI is not about replacing lawyers, but that "lawyers using AI will replace traditional lawyers." "There is a perception now, especially in professional fields, that if professionals like lawyers and accountants don't use AI, they will be replaced by people who use these tools because their work will be more efficient and they will be able to do more," he said. "

Inside the tech industry, one CTO noted that they started using AIGC to process service requests four years ago, prompting the company to now process about 89 percent of its unplanned service requests entirely autonomously. This front-end interaction has improved in responsiveness over time, and there has been no turnover in the team, the executive said.

"We're not laying people off, people aren't losing their jobs," the executive said. Human resources, operations, sales, facilities and legal staff in our organization are now using these tools. She also supports the above argument about the legal profession that those who use the ChatGPT feature have replaced those who do not use ChatGPT.

But another executive worries that the vision of people being freed up for higher-level jobs is too optimistic, and that the opposite may even happen in business, or that, as they put it,"people start to be challenged and stupid because we're used to turning to computers for everything."

One retail executive said current technologies, including robotic process automation and machine learning applications in prediction, are being used by "almost every industry," suggesting that more advanced AI isn't focused on replacing human jobs.

"It seems to me that almost every company has a strong desire to be more efficient," the executive explains. This is not from a labor perspective, but from an operational and financial perspective, so that customers get what they want most. What AIGC can do is help you compute, take your machine learning to the Nth level of limited levels, and it is far from reaching its full potential. AIGC is helping us find preset answers faster, in which case you don't have to train all models with your own data. "

In retail, this means AI can identify the best locations for retail stores and optimize the best routes for goods to be shipped to stores, but the key is that this process does not require the presence of data scientists at all times. If AIGC can free data scientists and machine learning experts from basic technical training and focus only on "increments," it will "greatly improve productivity," the retail executive said.

These smart machines don't require investment, they don't require expensive investment decisions, but they can help you understand potential outcomes and narrow them down to a few limited options rather than making blind decisions without a clue,"the executive said. And I think, for me, that's where the real potential lies. "

AI skeptics list a number of serious risks, such as AI has produced misinformation or inaccurate information, data noise (aka junk science) that could lead to costly dead ends for businesses, bias, threats to human jobs, and more. When humans communicate with AI, it is impossible to distinguish between the two. There are also copyright issues.

Technology has proven time and again what we can do without social barriers and technological barriers,"warns one executive. If you want to talk about the Industrial Revolution, let's also talk about how long it took to understand our impact on the environment. "

"It seems too late for us to think about privacy and put it back in the box," said another executive. "

A technical official in the consumer industry said that consumer product infringement problems may occur more frequently and there is no precedent to follow.

Recently, product development and technology have been integrated to create better experiences for customers, but now "we see our creatives using products like DALL-E and Midjourney to inspire their product development," the executive said. "He was referring to the very popular AIGC art generation program.

On the positive side, when they develop new products, it's a super great way to jump-start the creative process and do things like AI-driven mood boards. But on the downside, intellectual property ownership is harder to secure. After all, the technology is so powerful that it no longer needs to be provided through a professional technical organization, and people who don't have any technical experience can use it to improve what they're doing.

The news media is worried that misinformation will become more common, and academics are worried that students will cheat using new AI tools. But in business, another form of cheating is also seen as a big risk in technology adoption.

A consumer health industry leader says brands that start using AI indiscriminately could erode consumer trust in the technology when it leads conversations without greater transparency.

Recently, the American Journal of Men published an article written by AI, which provided inaccurate medical information. This could have a huge negative impact on the health field and raise questions about whether people know what they are reading is not human.

The executive said ChatGPT has already been tested in mental health apps and written by it, which could be useful. But there is also a problem with this process, that is, the relevant responses are not real people providing answers. To be clear, chatbot-powered mental health apps are not new in general, and many similar apps already exist.

But in a sense, getting robots close to humans is the AI goal proposed in the Turing test, that is, to reach a level where humans cannot distinguish between machines and humans in conversation. Moreover, companies will face serious challenges in disclosing AI responses and obtaining user consent,"which is difficult," the executive said.

But an executive at another company that has been using AI chatbots for years said the company has so far found that people love its service,"as we've seen, in many cases this orchestrated, personalised response interaction transcends human-to-human interaction, especially where there may be language barriers and other challenges."

Corporate boards need AI CEOs to use AI responsibly has always been a major part of the debate. One executive said company boards may need to create a position dedicated to AI. He added: "I can tell you this is going to push humanity forward, but it's going to take us slowly to figure it out. Just like management software, there are viruses and anti-virus software. "

This could also make AI a target for ESG investors, ensuring that companies use AIGC in an ethical manner. I mean, a company like nasdaq, a big investor like blackrock, they're definitely going to be looking at the use of AI,"the executive said." Just as you need ESG representation on the board, you need AI representation. "

The debate about AI will continue, but one executive who works with the logistics industry said that while adopting AI may bring many risks, it is equally inappropriate not to use AI to optimize processes for environmental impacts. AI-driven manufacturing and automation are critical to optimizing processes and yields.

"When I talk to our customers, I urge them to use AI to help address the environmental impact of waste and increase production," the executive said. Businesses should actively seek AI solutions because the technology is now powerful enough. "

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.