The latest stage of artificial intelligence evolution is here — and the outlook isn’t optimistic
The struggles faced by artificial intelligence startups have become apparent following the recruitment of Inflection AI cofounder Mustafa Suleyman by Microsoft CEO Satya Nadella. Ben Kriemann/Getty Images
- A new phase has commenced in the realm of generative AI, and it appears to be less rosy than its predecessor.
- Numerous AI startups are displaying signs of faltering after a period of significant optimism.
- Inflection AI experienced a setback as one of its cofounders joined Microsoft, while other startups are encountering challenges in revenue generation.
When the narrative of the generative-AI era of this decade is eventually chronicled, the recent month might be identified as the juncture at which the industry had to reassess the prevailing hype.
Since the introduction of ChatGPT, the buzz surrounding artificial intelligence has been pervasive. Companies with an AI-first approach have been driving the Nasdaq. Billions of dollars have been funneled into startups at an unprecedented rate. Visionaries like Bill Gates have likened the impact of this technology to that of mobile phones and the internet.
However, sustaining that euphoria has become more tenuous in light of recent events.
Startups that once secured substantial funding have tempered their ambitions. Ambitious founders who championed grand visions have now acknowledged the dominance of Big Tech. The notion that merely invoking “AI” could lead to lucrative outcomes appears less plausible.
In essence, a new era has dawned, characterized by a level of uncertainty in achieving success that was not initially envisaged.
Challenges Confronting AI Enterprises
A particularly noteworthy development in this new phase unfolded recently involving Inflection AI, a nascent competitor of OpenAI.
The company disclosed that its cofounder, Mustafa Suleyman, who also played a pivotal role in establishing DeepMind, was departing to lead a new AI division at Microsoft. Additionally, Inflection AI, which was valued at \(4 billion and had secured \)1.3 billion in funding as recently as June 2023 from prominent figures such as Gates, Nvidia, and former Google CEO Eric Schmidt, was deprioritizing Pi — its counterpart to ChatGPT — for consumer applications.
Mustafa Suleyman is poised to transition from Inflection AI to Microsoft.
Inflection
It is fair to say that this development sent ripples through the AI community. Inflection, after all, had garnered a valuation of $4 billion and had been exuding confidence publicly about its AI technology, which aimed to position its chatbot, Pi, as a “kind and supportive companion” for users. Suleyman even extolled personal AI as “the most transformational tool of our lifetimes.”
However, with its leader and a cadre of top engineers and researchers departing, those assertions now ring hollow.
Yet, it swiftly became evident that Inflection AI was not the sole entity grappling with challenges.
Stability AI, a unicorn startup behind the AI image generator Stable Diffusion, announced that its CEO, Emad Mostaque, was resigning. His rationale? Allegedly to pursue decentralized AI.
Meanwhile, The Information reported that another OpenAI competitor, Cohere, established in 2019 by former Google employees, only managed to generate $13 million in annualized revenue by the close of last year. The company is reportedly in discussions to secure additional funding.
Identifying the Missteps
Emad Mostaque is stepping down as the CEO of Stability AI.
Stability AI
These companies have been among the most prominent AI ventures of the past year. So, what led to their setbacks?
As noted by my colleague Ali Barr, AI faces challenges in the consumer domain. In the case of Inflection, its shift towards enterprise clientele coincided with Pi struggling to attract a daily user base exceeding 1 million. This figure pales in comparison to the vast consumer market.
While Inflection initially offered Pi as a free chatbot, there were contemplations about introducing a subscription-based model, as reported by Axios. The potential success of such a strategy remains uncertain in a landscape where competitors provide free-tier chatbots to entice the mass market.
Cohere opted to circumvent the consumer market entirely by focusing solely on enterprises from inception. Nonetheless, its struggle to bolster revenue underscores the unpredictable nature of buyer interest in enterprise-grade AI solutions.
On the other hand, Stability’s Mostaque seemingly conceded that Big Tech behemoths wield insurmountable influence in the realm of AI. In a post on X, he asserted that combating centralized AI necessitated a shift towards “decentralized AI.”
Not going to beat centralized AI with more centralized AI.
All in on #DecentralizedAI
Lots more 🔜 https://t.co/SbEF5zoo05
— Emad acc/acc (@EMostaque) March 23, 2024
However, few standalone entities appear capable of making significant strides in this domain without the support of established industry players.
Stability, which secured investments from private investment specialists like Coatue and Lightspeed Venture Partners, has encountered challenges in garnering the support of major tech firms that its competitors enjoy.
Should it aspire to progress into the “next growth phase,” merely identifying as an AI company may not suffice to ensure success in this evolving landscape.
Axel Springer, the parent company of Business Insider, has entered into a global agreement permitting OpenAI to train its models using content from its media brands.
