Mistral Sparks a New Wave of Controversial AI Investments

Recently, there's been a notable increase in announcements from various venture capitalists making minor investments in AI initiatives. Among these was the striking news of Mistral's $640 million Series B funding round, which has elevated its valuation to an impressive $6 billion.

To put it bluntly, these numbers seem unreal. Does anyone recall the ridiculous ransom demands in Austin Powers? It’s reminiscent of that absurdity.

Nevertheless, well done, VCs. You're playing directly into the hands of the dominant tech trio.

The Dominant Trio is Embracing AI

Now, here’s where I might stir some controversy: reckless speculation.

If you were tuned into Apple’s WWDC last week, you would have noticed their push to integrate Generative AI into their mobile phone empire, as they seem to do with everything else nowadays. Is the iPhone still the talk of the town? Imagine if it could make calls for you, even mimicking your voice.

The audience applauded. Casually.

Also, last week Amazon quietly announced a $230 million investment in Generative AI startups. While this might seem beneficial for new companies, I can't help but feel uneasy when one of the dominant trio claims to be assisting.

Here’s the crux of Amazon's investment: “A significant portion — including the entire amount allocated for the accelerator program — comprises compute credits for AWS infrastructure.”

This seems less about supporting startups and more about competing against Microsoft and Google.

On the topic of Google, they continue to be their usual selves, akin to an unrestrained 800-pound gorilla. They struggle with personalization, a crucial aspect of effective Generative AI, to the extent that I often think, “Please stop. You're so far off the mark, and I can't correct you.”

Oh, and why refer to the “Big 3” instead of the “Big 4”?

Meta—oh, dear Meta. You're developing something that nobody seems to want. I'm astonished by Facebook's decline over the past five years. The most telling phrase in recent discussions was a casual reference to data collected from the “adults of Facebook” (emphasis added).

Troubling.

It’s also worth mentioning that the Big 3 are heavily invested in the entertainment sector. Whether that holds significance is debatable. However, TikTok, which is not part of the Big 3, has utilized AI for years. Reflect on that algorithmic loop for a moment. This is what the Big 3's AI initiative is essentially inviting everyone to join.

What About the “Scrappy Second Tier”?

On one side, we have the corporate giants, and on the other, the emerging players.

OpenAI is not alone in wrestling with the ethical implications of AI, but it frequently finds itself embroiled in questionable practices, only to apologize after the fact—not after seeking permission, but after being denied permission by Scarlett Johansson for using her voice, then using it anyway.

If that's acceptable, then what’s with all the ethical concerns?

Since 2010, when I co-developed the first commercially available Natural Language Generation platform at Automated Insights, I've maintained that with such substantial investments at stake, “ethical AI” is merely performative, a distraction, a smokescreen. It merely opens the door to directing AI toward whatever business model one desires as long as it's labeled “ethical.”

Companies like Anthropic, xAI, and others are raising astounding amounts of capital during a time when the demand for Generative AI, though increasing, hasn't yet created a substantial set of use cases for mass adoption or significant traction.

Welcome to the Scene, Mistral

Now, let's examine Mistral’s recent move. I have no critique here other than that they are likely the last of the smaller players to make a massive leap, securing $640 million to bring their total to just over $750 million with a $6 billion valuation.

I won't criticize Mistral harshly. However, their Generative AI coding product is named “Codestral” and their chat product “Le Chat,” which suggests their marketing team is trying to downplay the inevitable costs associated with APIs in the future.

Furthermore, TechCrunch highlighted a statement from Mistral’s press release that concludes with, “It guarantees the company’s continued independence, which remains fully under the founders’ control.”

Immediately following that, TechCrunch mentions “General Catalyst.”

I’ll let that sink in. Also… troubling.

Next, they note that “General Catalyst led the Series B round,” followed by mentions of “Lightspeed,” “Andreessen Horowitz,” “Nvidia,” “Samsung,” and “Salesforce.”

I’m not criticizing Mistral.

Speaking of Nvidia, Amazon's investment in AI startups includes “access to expertise and technology from Nvidia, the program’s presenting partner. Participants will also be invited to join the Nvidia Inception program, which connects companies with potential investors and additional consulting resources.”

Does it not seem to you that “Nvidia Inception program” has a sinister tone?

Alright, perhaps I am being critical of Mistral. But connecting these dots is surprisingly straightforward. You just need to ask the right questions. Like…

What about the “ethics” of utilizing data from sources that haven’t opted in—or, according to their version of ethics, haven’t opted out yet?

What if they aim to disrupt the current SaaS and consumer markets by targeting input and output mechanisms?

What if their ultimate goal is to gather all the data?

This is where the agents in black suits storm my workspace.

A Refreshing Coffee with Jed

Now that I’ve unsettled you, let’s transition to a pleasant coffee chat I had last week with a colleague on a beautiful morning.

I mention this not just for a much-needed breather in my rant, but also because I’ve been critiquing all this hype surrounding Generative AI, and this discussion helped solidify my thoughts with someone knowledgeable enough to challenge me if I’m mistaken.

From 2010 to 2018, while I was involved in building and selling Automated Insights, Jed was running a different company in the same office complex. Our firms weren’t competitors, except in the race for skilled tech talent. We occasionally hired each other’s employees. Good times.

Jed’s concerns about AI mirrored mine, and as we discussed recent investments and corporate maneuvers, he concurred with my assessment.

He even added a crucial insight: “Whoever owns the data owns AI.”

So, here it is. You can blame Jed.

The Importance of Data Ownership

This brings us back to lessons I learned during my time at Automated Insights, the startup where we taught computers to generate articles from data.

True AI is incredibly complex. Generative AI is straightforward to initiate but challenging to execute effectively.

None of the Big 3 excel in either area, given their struggles with self-driving technology and the forgetfulness of their voice assistants.

I suspect this is intentional.

At Automated Insights, we demonstrated that possessing the right data to build models and create content is crucial. You don’t need an abundance of data, just the right type. It’s akin to an advanced game of “if-this-then-that.”

How does one acquire the right data? One way is to own all the data and identify the specific subset that is relevant at the opportune moment.

Well…

Amazon has access to all eCommerce data and has the infrastructure to process vast amounts of information.

Google remains the homepage of the internet and possesses the means to manipulate significant data volumes.

Apple is winning the battle to charge exorbitant fees for others to profit from mobile platforms. Just ask Epic Games about that. They’re located in my hometown, and on quiet nights, I can almost hear their frustrated outbursts.

Nvidia has the hardware capable of executing “if-this-then-that” tasks at lightning speed.

OpenAI, xAI, Anthropic, and now Mistral are all vying to gather as much data as possible.

So, how does one respond?

You can either collaborate with them or outpace them. This is why venture funding is either aligning with corporate behemoths or backing anything that exhibits an AI heartbeat. Yet, the latter approach seems futile and is slowing down. According to Crunchbase, early AI funding may be displaying signs of strain.

Thus, you are obliged to align with them. The robots aren’t taking over; we are the robots. WE ARE ALL PEOPLE!

Boom.

Everything is laid bare, and now I’m sweating and breathing heavily. If you crave more satirical insights in the future, sign up for my email list at joeprocopio.com. Perhaps you could also keep an eye out for my well-being when they come after me.