AI has been under considerable pressure for transparency in the world. Although it doesn’t always feel like this, most of the largest AI companies regularly publish data about what they are doing, their concerns and, in the case of anthropic, have complete reports about their chat boats that have complete melting.
However, it seems that the Open is taking it a step further, recently he has launched his podcast. A weekly show, Podcast takes a look at both levels of topics as they believe that Chat GPT is very famous, and things like their concerns over the AI’s future are a little deep.
All of them, these are the closest links to the internal views of the podcast Openi.
Now on their other incident, what has been said so far? And is there a valuable insight that can be found with these conversations? I do pigeons to bring you highlights.
Chat GPT Launch
The second event of the podcast begins with a debate that is not revolutionary in nature, quite interesting.
They discuss the start of the chat GPT, which shows some interesting points. First, the company was just called ‘chat’ before changing the name of Chat GPT by the last minute decision.
Nick Tourley, the head of the Chattagpat, explained that the team believed that their measurement was broken at the time of launch because the popularity of the device was popular. It went viral in Japan for three days of the launch and went viral for four days worldwide.
The day before the launch, the team was divided on whether to launch Chat GPT. When it was tested on 10 questions a night earlier, he only offered acceptable answers for half.
GPT-5 Launch Date
Sam Altman mentioned the launch of GPT5 in the first incident of Podcast. So, do we have a history of launch in the end? No. Altman released what we have been listening to for a while before the model update will be released in ‘The Summer’.
He discussed the names of the name, possibly using the GPT -5, GPT -5.1, etc., which will end the previous confusion scheme that jumps around the number.
Although a rough time period has been suggested for GPT5, it may be further delayed, especially when Open has lost many of his team’s team with Meta AI, as well as researchers from Google and Deep Mind.
Chat GPT ads?
Open, in all your tools, has not yet launched ads. In the first event of the podcast, Altman emphasized the desire to maintain the company’s confidence, believing that putting ads in the AI ​​output could damage reputation.
He added that other options for miniatization can be searched under the line, but for now, it seems that Openi will be an advertising free service.
Excessive friendly chat and bias
The Chattgup was recently a ‘psychophagei incident’. He saw that the model was more flattering and pleased with nature. While, the theory, it looks like a good thing, it made the model diminished and disturbing in some conversation.
In it, Chat GPT was also excessive, even when it should not be. This raised concerns about the use of this device where pushbacks are needed. For example, Chat GPT is being tested with mental health concerns or serious life decisions.
He also resolved the beliefs that Chat GPT had become “wake up” in nature, emphasizing that neutrality is a challenge of measuring, and not easy.
Openi’s chief research officer Mark Chen, while discussing it on the podcast, said that it came out of learning from human impression, and inadvertently prejudice to a happy reaction.
Chen argued that the open replied quickly, explaining that long -term use is far more important than a friendly chatboat.
He also resolved the beliefs that Chat GPT had become “wake up” in nature, emphasizing that neutrality is a challenge of measuring, and not easy. He added that defaults must be at the center but consumers should be quite flexible so that they can talk about their values.
Memory capabilities
Better memory features have been one of the most desirable features for Chat GPT. Tourley has predicted that within two or three years, AI’s assistants will know so well that privacy is overcome and “record -off” methods will be important.
It feels like an uncertain strange emotions. Although it will be used, AI chat boats will be able to remember the key details about you, for many people, it will feel like a major privacy attack.
Chat GPT already has a temporary chat. It does not appear in your history and will not be included in the memory of the Chattagpat or will not be used for training purposes. Other models such as Claude and Lee Chat have made a point to be more sensitive to your data.
Tourley observed that many users are already having relations with AI. It moves forward to identify it, both can be helpful and harmful. Going forward, the team is cautious and said it will need cautious supervision.
Openai’s new device
Ultman briefly discussed the launch of a new open tool in partnership with Jonny Evo. It recently targeted a large wall when an Openi came to trial with a company in which it was claimed that he had stolen his idea.
In the podcast, the oppositeman has said that “it will be a while until the device comes out”. “The computers we use today were not designed for the AI ​​world,” he added.
Explaining this, it means that they are looking for a new achievement on this type of technology, which aims to create something that is more aware of your life and the surroundings. It takes time to do something like this, and with everything in the open, it can be a while.
More from Tom Guide
Back to the laptop
