12/30/2023 – 9:00
If you had never heard the expression “artificial intelligence” before 2023, you certainly didn’t go unpunished during the year. After the success of OpenAI's GPT Chat, technology giants rushed to develop an AI tool to call their own.
“If we think that the end of 2022 was the launch of ChatGPT, the leap that was made with GPT4 was incredible. Certainly, 2023 was the year in which we saw an exponential advance”, pointed out technology expert Arthur Igreja.
+Fintechs are expected to grow double digits per year until 2028, says consultancy
News will continue in 2024
On December 6th, Google officially launched its multimodal language tool (LLM) called Gemini. The company promised that the platform is faster and more complete than GPT-4. Meta and Microsoft also announced their own tools. Igreja believes that big tech’s news in this regard will continue in 2024.
“Not only will it continue, but it will also increase in 2024, as it is on the agenda of all companies and has become something that is absolutely not only strategic but essential. It is a race for dominance, for applying artificial intelligence in the most varied technological segments. AI is a gigantic layer of technology,” he believes.
There is still room for improvement in the tools
After increasing the processing capacity of AIs, the expert believes that the next step to be taken will be to improve the tools, with better texts and more realistic images and videos. The issue of privacy will also be the subject of debate.
“Three years ago, videos created using AI were extremely precarious and now they are starting to take a certain shape. If there has been a very important advance in the world of static images, the next frontier will be video. Currently, AI still has many errors, especially in natural language models. A still significant rate of hallucination, lack of reliability in what is done. This will be refined. We are going to talk a lot about privacy, creating some kind of identifier for content that is purely synthetic, for content that has been tampered with by AI”, he points out.
Debate on regulation will advance
The European Union presented the first rules on the use of AI at the end of 2023, and this should encourage other countries to adopt similar measures. In Brazil, the debate on the use of deep fakes should take shape in a municipal election year.
“The challenges (of States) are intellectual property, copyright, veracity, reliability of these tools and misuse. Discussions on regulation will certainly advance. It will be necessary to know how to identify possible dangers that AI could trigger”, he concluded.
#expect #Artificial #Intelligence