But the experiments went further, we generated dozens of contents until the algorithm, within a focus on real estate marketing prospects in 2023, reported the following quote:
“Hong Kong Home Buyers Association President Mark Chien-hang told the South China Morning Post that although the value of newly completed homes in the territory is declining, the sales volume is still high. “This is because people with solid investment plans have already entered the market from abroad,” said Chien-hang.
This quote appeared as a element of reassurance fits perfectly into the context of the article. But before sending it online we decided to verify the source for the sake of completeness with the aim of citing it appropriately. Well, what we found out is that Mark Chien-hang doesn't exist, just as there is no Hong Kong Home Buyers Association. The cited text does not appear on the web and if someone were to think that there may be archives of information that feed these AIs that are not directly accessible online, well, that's not the case: the quote is completely invented!
Generative AIs have the ability to “generate” content in the style and form of the content they are trained on, but they do not “store” the data and are unable to restore it to its original form. Generative AI works like black boxes, made up of artificial neural networks to which no form of reverse engineering. In other words, the training data becomes numbers and there is no direct relationship between these numbers and the data that generated them.
To date they still do not find any commercial application due to the risk that they can provide completely incorrect information in sensitive contexts.
The OpenAI Terms of Use for ChatGPT read:
«Inputs and Outputs are to be considered collectively “Contents”. [...] The user is responsible for the Contents as regards the guarantee that they do not violate any article of the Law [...]»
If the information generated by ChatGPT exposes those who use it to the risk of publishing false, defamatory or other violations of the Law, for this very reason OpenAI protects itself by declining any responsibility for the disclosure of contents generated by its own artificial intelligence.
At the moment the great minds of the machine learning they have not yet been able to produce anything better than systems that answer every question only by lying through their teeth. And if these AIs sometimes respond correctly, it is a mere coincidence and nothing more.
Article of Gianfranco Fedele
Developing fine motor skills through coloring prepares children for more complex skills like writing. To color…
The naval sector is a true global economic power, which has navigated towards a 150 billion market...
Last Monday, the Financial Times announced a deal with OpenAI. FT licenses its world-class journalism…
Millions of people pay for streaming services, paying monthly subscription fees. It is common opinion that you…