Inflection received $1.3 billion in funding to develop “more personal AI,” the company said in June 2023. Microsoft served as the primary investor. Microsoft essentially stated now, less than a year later, that it was eating Inflection alive (although they phrased it differently).
Mustafa Suleyman, CEO of the newly formed Microsoft AI business, and Karén Simonyan, co-founder, will be joining “several members” (or “most of the staff,” according to Bloomberg) of their team at Microsoft. In an attempt to save what little remains of the company, which raised $1.3 billion nine months ago and another $225 million in mid-2022, Reid Hoffman will remain behind alongside new CEO Sean White.
The core idea behind Inflection was to create conversational AI that could be easily communicated across several platforms. It would remember your preferences and past discussions, enhancing the personalization and utility of the experience. Their AI, Pi, never even came close to implementing this wonderful notion. It was alright, but they still couldn’t compete with OpenAI, Google’s Gemini, and Anthropic despite their massive financial backing. Microsoft is also trading its bets with OpenAI.
Maybe the pitch was always going to fall flat; after all, it’s not obvious that consumers need individualized AI when these creatures are already so at home in the uncanny valley. Who knows? Maybe they don’t want their AIs for business writing, architectural sketching, and therapeutic discussion to interact with each other at all. Although unifying them is a viable strategy, we always considered Pi to be a credible runner-up, regardless of whether the market dismissed Inflection’s tech or it failed to achieve its goals.
Taking this shot was expensive, like tens of millions of dollars, and probably didn’t bring in much money. You keep pouring money into the hole, but for how much longer? At what point is it no longer a long play and becomes a capital loss? To top it all off, who decides?
Regardless of who it was, “The New Inflection” is currently available just without its two most technically significant co-founders, without the product in which it had invested heavily, without part or all of the development team, and one can only assume, without a significant portion of the funds that were tied to performance. It will be difficult for Hoffman and White.
A year or two ago, the new emphasis on an “AI studio business, where custom generative AI models are crafted, tested, and fine-tuned for commercial clients” may have been acceptable. However, they might now be competing for bits.
Lastly, we have Microsoft. Are they saving a valued team from the ruins of a company that might have brought them down with it, or are they the villains in this situation? On the other hand, are they sneaky people who bet on many horses in a single race just to feed themselves on the one that fell before the finish line?
Seriously, we didn’t see them attack OpenAI until November! Sam Altman, Greg Brockman, and their devoted followers were the meat, while OpenAI served as the fallen horse for a few hours! Instead of devouring the company, Microsoft managed to gain some additional leverage.
Their second option is now available to them and will likely be far more affordable. When OpenAI was at its most ambitious before the coup, it had a lot of momentum, which Inflection lacks. Suleyman and Simonyan’s ability to successfully head the new AI branch will be tested. After witnessing the rapid demise of a billion-dollar venture, we must admit that we are rather doubtful of their intuition. Will you still employ them after this?
Regardless of the causes, this unexpected decline highlights the dominance of established tech businesses in this field. Microsoft was giving them money and computers, whether it was OpenAI or Inflection. They talked to them about partnerships, and when they tripped, the secret fork and knife came out.
Do you believe Google isn’t prepared to proceed with the same approach towards Anthropic? Why would Apple not follow suit if given the opportunity? The tech industry is notoriously cutthroat: the strong create, while the weak acquire whatever is needed.
