Billionaire tycoon Mark Cuban has a theory on what Elon Musk’s long-term plans are for Twitter, and they have precious little to do with protecting civil rights enshrined in the American constitution.
“He isn’t pushing free speech,” the Dallas Mavericks owner and Shark Tank star shared on Twitter on Thursday. “When Elon Musk goes out of his way to engage with and promote the accounts and tweets of people who disagree with him, then I’ll believe he is about free speech.”
Nor did the investor known for his enthusiasm over cryptocurrencies agree with Musk biographer Walter Isaacson that the big goal for the once-ousted PayPal CEO is disrupting the financial services industry.
Instead, Cuban believes Musk is primarily focused on developing his own version of ChatGPT after the visionary entrepreneur’s very public break with its creator, OpenAI, which he co-founded in late 2015 and helped finance.
“He gets to take the entire Twitter firehose to train or feed any open-source model and have a competitor to the Big 3,” Cuban continued, referring to Google, Meta and OpenAI together with its partner Microsoft.
Musk believes he was instrumental in recruiting OpenAI’s key talent, first and foremost chief scientist Ilya Sutskever, that ultimately stole Google DeepMind’s thunder.
All of this is why I think @elonmusk is starting his own @truthgpt, or whatever he will call it. He gets to take the entire @twitter firehose to train or feed any open source model and have a competitor to the big 3. He can weight his own tweets and those of the sources he… https://t.co/tWk4Kfp8tv— Mark Cuban (@mcuban) May 18, 2023
Yet he no longer holds any formal role after the company traded in its non-profit status and stopped publishing all its proprietary research.
Musk likened it this week to funding an “organization to save the Amazon rainforest—and instead they became a lumber company, chopped down the forest and sold it for money.”
But Musk has the contacts and the plans to create what he has thus far dubbed TruthGPT.
Through the application of large data sets, for example, those generated by the millions of Twitter users every day, a GPT can learn over time to improve its own natural language response data.
“He can weight his own tweets and those of the sources he likes and end up with a consumer-facing AI that can be a virtual Elon,” Cuban wrote. “Pretty cool. Pretty scary.”
Not the time to show his cards on TruthGPT
Generative A.I. has become the hot topic ever since OpenAI unveiled ChatGPT, the fastest-growing application in history since it launched on Nov. 30th. Goldman Sachs has argued it could increase profits at S&P 500 corporations by 30% or more while Cathie Wood’s ARK Invest believes it could quadruple the productivity of knowledge workers.
Now OpenAI is bringing the power of its advanced chatbot to Apple smartphones, announcing it would bring an app for iOS and soon even for Android, the proprietary operating system of its A.I. rival Google.
Yet experts have warned that risks are not limited only to presenting incorrect statements as facts, which prompted the MIT Technology Review to call A.I. chatbots “notorious bullshitters” earlier this year.
What we need is TruthGPT— Elon Musk (@elonmusk) February 17, 2023
Instead, they could be weaponized to become agents of misinformation, according to the European Union’s disinformation watchdog, the European Digital Media Observatory.
EDMO argues those users not particularly well informed about the technology’s strengths and weaknesses could be tricked into giving more trust, or relying more heavily upon, the results provided by a conversational chatbot than other more open-ended tools.
This makes it more difficult “to properly assess the veracity of the results, and ultimately to exert their critical thinking,” it wrote last month.
Speaking earlier this week, Microsoft CEO Satya Nadella agreed that A.I. deployed by governments such as Russia—or indeed non-state actors—could manipulate voter behavior in the run-up to next year’s U.S. Presidential election.
“Does A.I. further accelerate and give tools, absolutely it does,” he said in an interview.
A chatbot created by Musk could in theory skew responses against ideas supported by the Biden administration and the Democratic party, of whom Musk is a vocal critic.
When the Twitter owner, who openly called for voters to cast their ballot for the Republican party, was asked about his plans for TruthGPT, Musk declined this week to explain his vision for the project.
“We will have a launch event and we will explore the issues in more detail,” he pledged. “This is not the time.”