Tesla is also heavily researching self-driving technology, but now it’s very focused on the Ai side, too. It’s going to show us growing faster and faster after some time, right? However, it still seems like a lot of data collection is needed. Let’s complain about a sudden stop where a Tesla user exceeded the FSD speed limit. “The sudden stop over the FSD speed limit is kind of crazy,” Elon Musk responded.
12/15/2023 7:38pm. (Benzinga Newswire)
Tesla CEO Elon Musk acknowledged Twitter users’ concerns after users complained on X (formerly Twitter). Users complained that Tesla cars would stop working if they exceeded 85 mph in Full Self-Driving (FSD) mode.
Last Friday, an FSD user spoke to X and said the immediate stop when crossing the 85mph speed limit was “a little crazy.”
His claim is that even in severe situations, even if it exceeds 85 for a moment, it will disable FSD. The user has suggested a potential solution as well.
He recommends a 10-second timer instead. According to the idea, Tesla drivers should disable FSD for 10 seconds after receiving a warning or go back below 85.
Musk simply responded to an avid fan’s suggestion by saying, “Ok.” The CEO did not elaborate further.
Musk has been teasing FSD version 12 for months. In June, Musk showed signs that version 12 will no longer be a beta version, but will achieve full self-driving.
The current FSD is in beta mode, still being tested and requires active driver supervision.
“We’re very close to the full autonomous driving capability,” Musk said at the 2023 World Artificial Intelligence Conference in Shanghai in July.
“I’ve been wrong about this prediction in the past, but I feel like we’re closer to it,” Musk said. Elon Musk’s Grok poses a serious threat to ChatGPT
Musk’s startup xAI described the chatbot as “toxic” and ready to “answer spicy questions.” Only a select group of beta users had access.
Over the past week, however, Musk has significantly expanded access to the record by making it available to anyone who pays $16 a month for a premium subscription on X.
After a few days of trying, I found out that Groc was better where some competitors were left behind: a real-time summary of the news.
I asked him about three new topics: the international climate change negotiations, the Israel-Hamas war, and the licensing agreements OpenAI has with publishers
The book answered with concise and largely accurate answers, although there are some questionable sources.
Nevertheless, in a limited part of my experiments, it performed better than the Bard and ChatGPT operating on paid GPT-4 models with access to real-time data.
Grocery was the only chatbot I tested, quoting articles from Bloomberg and the Financial Times that correctly stated that Saudi Arabia agreed to a new goal at the recent COP28 climate summit.
In contrast, ChatGPT and Google’s Bard said Saudi Arabia did not agree to cut emissions, citing outdated news reports earlier this week that reportedly stalled negotiations.
When I asked for an update on the Israel-Hamas war, Brock said that the conflict had “entered a new phase.” According to the Gaza Ministry of Health and the IDF, Brock gave an estimate of the death toll.
But Bard refused to get involved in the Israel-Hamas. “I can’t help because I’m just a language model and I’m not capable of processing and understanding it,” Bard said.
And when I asked OpenAI if it was paying publishers for content, he was more accurate than OpenAI, and that morning he heard the news that OpenAI had signed Axel Springer.
Known for co-founding OpenAI at the time, Musk chose xAI as his rival ChatGPT. His company now expects to raise $1 billion.
But the biggest advantage of Grocery is that it can draw social media posts from X’s API, which is a repository of real-time information
Musk has previously cracked down on other companies accessing the same data.
Not everyone is going to pay a subscription fee to use Groc. And by xAI’s own admission, Groc isn’t as advanced as the latest models from industry peers like OpenAI and Google.
But according to some inference and knowledge benchmarks, the company says its most advanced model, the Groc-1, outperforms OpenAI’s previous GPT-3.5 model and Meta’s popular open-source llama 2.
The lock was announced four months after Musk announced the creation of xAI. OpenAI and Google worked on iterations of their current models for many years. That will make OpenAI worry.