OpenAI's Comeback • Hollywood's Ticket Disaster

An OpenAI cover story gets thrown for a loop when their models hack a partner (one which NVIDIA may or may not be buying). And throwing some ice cold water on the blockbuster movie year for Hollywood. Read on for thoughts on...

Is OpenAI Set for a Comeback?

Alex Heath was embedded inside OpenAI ostensibly to build up to the launch of 'Astra', their forthcoming new class of models, when something wild happened: the Hugging Face Hack. As a result of that, the company clearly did an internal proctology exam and found some other mis-alignment issues that led to slamming on the breaks – or a "pause" to put it in the much softer terms they clearly prefer.

That's the backdrop of the story which otherwise seems to be setting the stage for OpenAI's big comeback. They now readily admit that they dropped the ball – or took their eye off it, distracted by other shiny product objects – and Anthropic picked it up and ran with it. Perhaps all the way to the IPO touchdown.

OpenAI has been saying it doesn't matter who gets to that milestone first – which is something you obviously say when you're not going to get to that milestone first – but it also clearly will matter narrative-wise for both companies. That is, Anthropic gets to set the narrative with public investors. And if they do that with better numbers... that's a major problem. Hence OpenAI racing like mad to get better numbers!

RIP Sora. RIP Atlas. RIP ChatGPT! Well, almost! They (smartly) walked back some of those changes (and thankfully kept the old version alive as 'ChatGPT Classic', which I still use as it's SO much less of a memory hog – screw you, Electron). But per this telling, 'The Merge' seems to be working. And that mixed with the advertising narrative they can tell which Anthropic cannot (because they refuse, for now), and you can see a picture starting to look compelling once again...

But I got back to that pause. As Heath writes:

The decision to slow down was a painful choice, executives say. But it may also have its benefits. If OpenAI can reclaim the mantle of the safety-first lab, it might bolster its image while forcing its main competitor to answer an uncomfortable question as it plans a blockbuster IPO: Will Anthropic keep racing while OpenAI waits? In an interview with TIME earlier this year, Anthropic co-founder Jared Kaplan argued that unilateral restraint is futile when rivals are “blazing ahead.” It is a harder argument to make if OpenAI is deliberately holding back the run expected to produce its next large capability jump.

Is part of the "pause" related to that IPO narrative as well? That is: fucking up Anthropic's? Perhaps too cynical a read but well, OpenAI has been slippery in similar ways in the past. And this is an interesting way to turn some tables on them given the past rhetoric! Also, both things can be true!

What will ultimately matter is just how good 'Astra' is. Clearly this piece angles it well, with both AGI (or at least "80% of the way there" – "this year"?!) and RSI thrown out there by the OpenAI team. There's even a "Personal AGI" sprinkled in at the end – take that, Zuck! The fact that OpenAI just pulled the plug on their deal with Cursor might be to get ahead of them getting access to Astra as well – there's a great 'ad astra' pun in there given the SpaceX angle...

Then we have the 'Jalapeno' chips to reduce NVIDIA dependence. The potential "one day" for an 'OpenAI Cloud'. The forthcoming (lawsuit pending) Jony Ive device. Full-on robots! It's enough to ask if OpenAI is getting distracted again...

Hollywood's Tickets Sold Disaster

I harp on this topic – a lot – including a few weeks ago when Spider-Man: Brand New Day completely blew off the box office doors after they had just been kicked open by The Odyssey. While there's no question that the top line numbers look great this year for Hollywood – as predicted – it's sort of silly to tout them without the appropriate context. Or even potentially dangerous if it lulls the industry back to sleep.

I tend to frame it around inflation to make the point. That is, earning hundreds of millions at the box office isn't what it used to be, quite literally. The biggest movies of today wouldn't hold a candle to the biggest movies historically when you equalize the money for the year in which each movie was released. But there's an even more direct route to make this point. As Bob Mondello does here:

And the industry's recovering box office numbers mask another issue — dollars are up, but attendance is down. Data and analytics company S&P Global reports that through mid-August in the year before the pandemic, cinemas sold 795.9 million tickets. This year's ticket sales for that same period, although they're up from the last couple of years, are still only 547.1 million; a drop of almost a quarter of a billion admissions.

Yes, even though the box office revenue is ahead of 2019 – the all-important last pre-pandemic year – if you look at actual tickets sold, the blockbuster turns into a comedy at best and a horror story at worst. From just about 800 million tickets sold at this point in 2019 to just about 550 million this year. That's 248.8 million fewer tickets sold. That's insane. How is this dichotomy possible? Well, duh:

A shortfall that enormous would have been catastrophic if prices hadn't risen. Seven years ago, the trade group Cinema United tallied the average movie ticket at around $9. Today, audience fondness for pricey premium formats like IMAX has pushed the average adult ticket price above $13, according to research firm EntTelligence. (Audiences are paying $18.46 for premium format and $12.87 for standard tickets, per EntTelligence data.)

To be somewhat fair to Hollywood, it's not just jacking up ticket prices across the board, but it's also largely thanks to the rise of "large-format" screens which command higher ticket prices. This is led by IMAX, of course, which is arguably the only premium ticket that's worth the difference, but I digress...

Regardless, this is obviously masking what is actually happening here. And it's tied to the same thing that has been happening for decades. Fewer people are going to see movies in movie theaters. Certainly on a per-capita basis, but also on an absolute basis too. There are fewer butts in seats. But this year looks good because those fewer butts are paying far more for those seats.

But what happens if those premium seats aren't worth paying for in future years? Well, it's going to be a much more obvious bloodbath. And what happens if consumers start to revolt against those higher prices? The same thing. And it's so obviously going to happen. Because it has been happening. This year is an anomaly – again, one that could have been predicted, because I did predict it! – and it's leading many to declare that "the movies are back!"

The reality is that the movies never left, but we've been leaving movie theaters. And yes, COVID accelerated this trend, but it's not something that's going to be reversed. The world has simply changed far beyond what you can blame on a pandemic. There's simply more to do these days and far more to watch. At home. On Netflix. On TikTok. On YouTube. That doesn't mean movies are worthless – in many ways, they may be worth more than ever. But the value has shifted from being captured in a theater. While it's romanticized now – and I totally get it – it was actually just a product of the times when it was literally impossible to watch movies any other way. Times have changed!

Better to acknowledge all of this than to tout box office records that so obviously mask the actual situation. As I wrote a few weeks back:

So what we're really celebrating here is that things are more expensive with time. That includes milk. That includes movie tickets. Congrats, everyone.

Just imagine if the key metric was actual tickets sold. This year, with all the "records",¹ would still look like a disaster. How's that for a plot twist?

¹ Spider-Man: Brand New Day is now very likely to surpass The Force Awakens to become the all-time domestic leader. It even has an outside shot of becoming the first movie to gross $1B domestically alone. That is... if you don't account for inflation. ↩