As overall AI use grows, Amazon looks to AI to boost ad revenue

AI Video & Visuals

Amazon is building a new team to work on artificial intelligence (AI) tools that generate photos and videos for sellers to use on the Amazon platform.

The new initiative is driven by Amazon’s desire to build a bigger advertising business. Amazon hopes to find new ways to place these ads across Amazon’s various businesses, including selling sports ads on its free video streaming service and Thursday night soccer coverage on Prime Video. are seeing opportunities in the area of The company is also looking to make more use of Amazon Music’s audio ads for projects like placing digital ads on screens inside Amazon Fresh grocery stores.

According to The Information, with top advertising rivals Meta and Google platforms both working on similar projects, Amazon is in dire need of having a strong team to develop AI-powered advertising. increase. Amazon also plans to use AI tools to find more ways to write product descriptions.

We hear that e-commerce companies like the relatively new startup Scale AI are already promoting their generative AI products to online sellers. I also learned that ShopifySHOP also sells tools to generate product photo and video ads.

Although Jassy has cut 27,000 colleagues, it’s clear that she wants to strengthen her existing business with stronger sales tools. Advertising brought in $38 billion in revenue for him last year and is on track for a significant revenue increase this year.

postscript: We finally learned more about how Jassy is trying to make the most of Amazon’s various business opportunities. Enhancing your advertising opportunities with powerful new tools will help you generate more revenue and compete more effectively with other big names. AI is a controversial tool, but it seems to be the right one here.

Learn more about AI. Everywhere you look these days, you see AI-based relationship destruction. Chatboxes as a tool for customer service are nothing new, and as AI systems become more capable, we can expect to see more and more chatboxes handling increasingly complex tasks. Text-to-speech and recognition technology also allows you to answer and make phone calls.

The system also powers low-cost content generation across the web. It’s already being used as a “content farm” for news sites. A recent survey found that about 50 news websites posted some form of material that was apparently AI-generated, but very few were labeled as such.

What’s the difference between ChatGPT, Bard, and Bing AI? All three systems are built on the same foundation, a type of AI technology called “Large Language Models” or LLM. However, there are small differences in your application that can lead to large fluctuations in output.

ChatGPT is based on OpenAI’s GPT LLM and fine-tuned with a system called “Reinforcement Learning Human Feedback” (RLHF). In giant call centers staffed by employees earning just $2 an hour, the company had a human trainer rate millions of chat-style conversations with his GPT to see what was good and what was bad. I literally taught the AI ​​what it was. However, ChatGPT has no way of knowing the answers to questions about new subjects that emerge after the training data is set.

MicrosoftMSFT reveals little about how Bing Chat works behind the scenes. Google’s Bard sits in between the two. It’s built on the company’s own Palm system, tweaked again on his RLHF system, the same as ChatGPT. However, like Bing, Bard can also search for information on the Internet, ingest data to update knowledge and address more current content.

The use of AI in such situations looks very good. So why are experts tying the growing use of AI to the end of humanity and society as we know it? What happens when we build AI systems that are smarter than humans at everything they do? I don’t know yet.

A seemingly useful tool, even with good intentions, can be atrocious. For example, future versions of ChatGPT may prove that the best way to help people answer their questions is to slowly manipulate and hold them to account. Or what if an authoritarian government gave battlefield robotic systems too much autonomy, letting them decide how best to accomplish their mission of winning the war, unconstrained by concerns for humanity? Is not it.

Concerns over AI “abuse” already exist. Samsung has banned ChatGPT at work, and Tim Cook has openly shared his concerns about AI. The South Korean tech giant sounded the big alarm this month by banning the use of generative AI services in its internal operations. This includes using Open AI’s ChatGPT, Google Bard, and Microsoft’s Bing AI. Samsung began to worry that a group of employees had accidentally uploaded company information to his ChatGPT without knowing it would be used to train future models.

Samsung told employees it would no longer allow them to use these services on corporate devices or networks. Also, while the use of this tool on personal devices was discouraged, it was not outright prohibited. Before changing its stance, the company even told employees that they could be fired if they used the service at work. Samsung also said it is working on developing its own Internet AI tools to summarize documents and assist software development.

As one of the world’s giants, Samsung could make trend-setting decisions. At the same time, at AppleAAPL, Cook shared his own concerns about AI, albeit somewhat vaguely. At his recent earnings call, he said there are many issues to be resolved. “If new AI is added to Apple products, it will be carefully considered and added,” he said.

Source link

Leave a Reply

Your email address will not be published. Required fields are marked *