Hey Everyone,
With rumours swirling that Charcter.AI is seeking several hundred millions of dollars of funding from Google, OpenAI recently told the Financial Times that it’s seeking a lot more funding from Microsoft for AGI development.
Good sources for these sorts of rumor/breaking scoops are The Information, Semafor, Reuters, FT and occasionally Bloomberg.
As Reuters noted, Character.AI’s character-based chatbots have struck a chord with younger users who make up 60% of its web traffic, helping position the company as the maker of a more “fun” AI companion when put up against ChatGPT or even Google’s Bard.
ChatGPT is not viewed as entertaining and hence struggles to convert daily active users from monthly or “weekly active users”. It claimed to have 100 million weekly active users on its first DevDay.
Character.AI allowed people to create their own chatbots a long time before OpenAI came out with GPTs, sort of like a lesser powered clone. Furthermore OpenAI does a lot of marketing that it’s leading to AGI, which probably isn’t based on fact.
I’m not a paying subscribe for fluff like The Information or FT PR pieces, but here is the article in question. Various outlets have unpacked it for us. When Anthropic was able to secure funding from both Google and Amazon, it must have upset the apple cart at OpenAI and they are seeking more funding from Microsoft, perhaps Softbank and rumored also the Middle East. They would likely want Billions for their hype-ship to develop AGI.
Altman said his company’s partnership with Microsoft and its CEO Satya Nadella was “working really well” and that he expected “to raise a lot more over time” from that company and other investors to fund the cost of creating more sophisticated AI models. Well Sayta is certainly going to hype it since Microsoft gets 75% of the profits.
More Funding from Microsoft means OpenAI loses its Majority Stake
More funding from Microsoft would mean they would own more than the 49% they will end up owning. Microsoft will reportedly get a 75% share of OpenAI’s profits until it makes back the money on its investment, after which the company would assume a 49% stake in OpenAI. I can no longer find the original Semafor article that breaks down the deal, for some reason.
The hunt for AGI is more or less a fiction Silicon Valley has made up. OpenAI is exploiting this for marketing purposes.
Earlier this year, Microsoft announced a $10 billion multiyear investment in OpenAI, adding to a previous $3 billion investment and, giving it a 49% in the company which was then valued at around $29 billion. Now OpenAI claims its worth $86 Billion.
At $86 billion, it would leapfrog the likes of Stripe and Chinese online retailer Shein to become one of the world’s most valuable closely held companies, behind Elon Musk’s SpaceX and TikTok parent ByteDance. The idea that OpenAI is worth almost as much as SpaceX doesn’t make much sense.
But the Generative A.I. golden goose is transforming VC interest in Generative A.I. into a very weak rocketship and startups Anthropic and others are cashing in. OpenAI wants to raise likely Billions more. Even ElevenLabs is due for a big one with a16z and Sequoia soon. The rumors are really swirly as we head to the end of 2023.
OpenAI is also working on GPT-5, the next generation of its AI model, Altman said, although he did not commit to a timeline for its release and Silicon Valley executives are saying AGI is in the works and on the horizon.
Since its initial launch in September 2022, 20 million users have signed up for Character.AI, though traffic to its website has been declining since its May 2023 peak, per analytics provider SimilarWeb. OpenAI claims ChatGPT has 5x that, at least as convenient “weekly active users”. It’s not clear if GPTs will be fun, but it doesn’t appear like they will.
OpenAI pumping up “AGI” as a realistic goal might have been Microsoft’s idea all along, who knows. That Microsoft Researchers were talking about “Sparks of AGI” even with early access to GPT-4 was quite concerning. It was a major red flag for me that this is nothing like the AGI we were told about growing up.
By changing the definition of what AGI is, Open and Microsoft plan to scam the world and it’s truly quite a Venture Capital and Enterprise level FOMO spectacle. I’ve never witnessed so much misinformation about something online sponsored by Silicon Valley before.
In an interview with the FT, Sam Altman said the Microsoft partnership would ensure “that we both make money on each other’s success, and everybody is happy”.
OpenAI is worse than a vassal of Microsoft, it’s practically its R&D and AGI hype machine. Though it is more commercially driven than DeepMind ever was for Google (Alphabet).
While GPT-3.5 was amazing compared to what we had before, GPT-4 was not in of itself mind-shattering like the media has claimed. That the “winners” of this mini-trend of May 2022 to December, 2023 are raising even more funds into the hundreds of millions into the multiple Billions, means the trend won’t be able to sustain its own hype for every long. People are betting on something that isn’t as transformational as is being claimed.
There’s no real evidence that GPT-5 would be “closer to AGI” or significantly a huge improvement like we saw between GPT-2 and GPT-3.5. But time will tell, Microsoft has already cashed in with Copilot, GitHub Copilot, Bing AI and various other schemes. Another 5-10 Billion from Microsoft with Softbank adding 1 Billion and the middle East adding another 2 Billion is well within the limits of this hype train. The higher valuation of OpenAI internally, implies an ability to entertain the prospect of unheard of funding rounds from here on out.
However, despite growth in revenues OpenAI remains unprofitable, in part due to high model training costs. It becomes the ultimate cash burning machine due to the cost of compute and cost of talent. It’s not as if ChatGPT is that useful to the common person, inspite of influencers telling you otherwise, even professors who conduct studies paid by the partners of OpenAI themselves!
We don’t even know how many of these AI startups will survive the 2020s, much less ever be profitable.
By asking for more and more funding in an era where interest rates are higher than ever, it’s not necessarily the best thing or sustainable. By the pressure for more funding and faster growth has never been higher for the leading A.I. startups.
If ChatGPT was really a tool approaching AGI, it wouldn’t just have 2 percent of Google’s traffic. Google Search has been “relatively stable” over the past 12 months with Bing disappointing Microsoft executives with forced adoption of Edge not going so well.
If OpenAI was able to get say 20 Billion more funding in early 2024, it’s not clear what would happen to society or to the pursuit of AGI. There’s no reason to actually believe we are anywhere near genuine AGI. With Nvidia’s new chip announced, OpenAI would have the ability to out-compute its rivals with more significant funding though.
For Azure, it’s always been just a game for more growth. Does Microsoft really care about AGI or does Google? Only in so far as it pads their profits and marketshare dominance, enables new revenue streams and strengthens their already anti-competitive moats.
Character.AI, OpenAI and ElevenLabs among others, should be able to get more funding soon. All bets are on, even in the aura of AGI marketing.
Read More in AI Supremacy
