What OpenAI’s wave of releases says about 2024 | The AI Beat

OpenAI’s latest release announcements during the company’s first developer conference, Dev Day — custom GPTs! New GPT-4 Turbo! Assistants API! — crashed over Silicon Valley and the world today like a massive wave of hype and excitement that threatened to sweep us all away with the current.

Of course, observers of all things AI, like me, were already soaked to the skin, exhausted by last week’s tsunami of newsworthy AI announcements.

There was the White House’s AI Executive Order. The G7’s voluntary code of conduct. The UK Safety Summit. An amped-up x-risk debate by the ‘godfathers’ of AI; a ruling that pared down a prominent AI copyright case; and updates from Midjourney, Runway and Stability AI.

But wait, there’s more: Scarlett Johannson takes legal action against AI app! Google rolls out GenAI tools for advertiser product images! AMD soars on AI chip sales predictions! Collins Dictionary selects ‘AI’ as the word of the year!

And as usual, the weekend brought no rest for the AI weary, as Elon Musk released xAI’s first LLM, Grok, on Saturday.

The timing, therefore, could not have been better for Sam Altman to announce a slew of new capabilities and pricing changes for its AI platform — Cheaper! Better! Heading towards AGI! After all, with just a few short weeks to go before ChatGPT enjoys its first birthday, OpenAI was due for a giant wave that we all have to figure out how to ride. Surf’s up, dude!

Predicting the AI tides of 2024

After a year and a half at VentureBeat, I’ve come to believe that every week in the wild world of AI news offers some clear takeaways. After the past week’s wave of announcements, it feels like some in the AI community are racing towards 2024 with a rising tide of hope (or in some cases, hubris) and, in some cases, a flood of fear.

These are three key ways I think this latest wave of AI news signals trends for 2024:

1. AI pause, shmause.

Seven months ago, Elon Musk signed an open letter calling for an ‘pause’ on large-scale AI development beyond GPT-4. This weekend, he did just the opposite — he debuted xAI’s Grok, an LLM offering realtime data from X and a sarcastic bent, releasing it to “a limited number of users in the United States.” 

Whether or not you thought that a six-month pause was a useful or doable idea, it’s clear that for Musk, as well as OpenAI and other LLM companies, there will be no pause. If anything, 2024 will be the year that AI development moves to hyper-speed, especially as enterprise companies get closer to being able to put AI use cases into production (and China makes its own AI money moves at a fast pace).

2. Regulators will be working overtime.

Today’s OpenAI releases came in the wake of a busy week of AI regulation announcements. But the mic drop of all things GPT proves that this is just the beginning of the challenges for regulators: For example, President Biden’s ambitious AI Executive Order may signal an effort to keep up with AI driving change at “warp speed,” but the truth is, it is just an initial step that will need to be augmented with congressional legislation that will likely be discussed, negotiated and debated throughout 2024.

The UK Safety Summit, too, was a jumping-off point for 2024 — apparently South Korea will hold a second Safety Summit in six months, while France will host a third as 2024 draws to a close. And while some say there is still a 50-50 chance the finalized EU AI Act could see the light of day by the end of 2023 (apparently there are more than 100 lines of text that are not yet agreed upon), the fact is that the the act would not be adopted until at least mid 2024 before the European Parliament elections. And, how those regulations would be enforced remains to be seen.

3. Enterprises protected on copyright issues — not consumers.

During today’s OpenAI announcements, CEO Sam Altman called out the company’s new “Copyright Shield” — which means, he explained, “that we will step in and defend our customers and pay the costs incurred if you face legal claims or on copyright infringement.”

By “our customers,” he did not mean the individual consumers buying a monthly subscription to ChatGPT; he was talking to the developers at enterprise business customers building on OpenAI’s APIs. OpenAI is far from alone in this — Microsoft, IBM, Shutterstock and Adobe have already made similar promises, while Google Cloud made its own announcement a couple of weeks ago, telling its business customers that “if you are challenged on copyright grounds, we will assume responsibility for the potential legal risks involved.”

This is a trend that will surely only expand in 2024, as generative AI becomes ever-more integrated into products; as the number of copyright-related lawsuits continue to rise, and as LLM companies and cloud providers continue to look for ways to attract enterprise business customers that want to protect themselves and limit their liability.

Get ready to ride the AI waves of 2024

With only eight weeks to go until New Year’s, there are still likely many more AI announcements to come before the briefest of pauses around the December holidays.

So towel off. Shake your legs out. Wax your surfboard. Get ready to ride this year’s AI waves right into 2024. I’m stoked — and hope to be hanging ten with you all.