Greetings,
The AI hype is hitting a wall of reality.
While the funding frenzy continues, new data from S&P Global shows a surprising trend: 42% of companies abandoned their AI pilot projects in 2024, a massive jump from just 17% the previous year.
The data suggests a growing chasm between launching an AI experiment and proving a clear, scalable return on investment.
Many firms are getting stuck, unable to move from flashy demos to full production deployment due to complexity, cost, and talent shortages.
This trend is creating a clear split between the AI haves and have-nots—companies that can successfully integrate AI and those that will be left behind.
But for the companies that are making it work, the scale is staggering.
Some key data bites from this week that you should know:
OpenAI employees plan to sell $6 billion in shares in a secondary deal valuing the company at $500 billion.
Anthropic is in talks to raise $5 billion in a new round at a massive $170 billion valuation.
AI search engine Perplexity made an unsolicited $34.5 billion cash offer for Google’s Chrome browser.
Cognition secured nearly $500 million in new funding, pushing its valuation to $9.8 billion.
Foundational model builder Cohere raised $500 million in a new round, valuing the company at $6.8 billion.
The AI industry has now produced 498 unicorn companies with a combined valuation of $2.7 trillion.
Google is investing an additional $9 billion to expand its cloud and AI infrastructure in Oklahoma.
Nvidia-rival Cerebras Systems is in talks to raise approximately $1 billion in a private funding round.
The market for AI “companion” apps is on track to exceed $120 million in revenue in 2025, with 17% of active apps including the word "girlfriend" in their name.
Perplexity is reportedly seeking another new funding round that would value the company at $20 billion.
A new hedge fund run by a 23-year-old former OpenAI researcher is now managing $1.5 billion in assets and delivered a 47% return in the first half of 2025.
OpenAI CEO Sam Altman is reportedly co-founding a Neuralink competitor named Merge Labs, which is raising funds at an $850 million valuation.
In today's newsletter:
💰 Funding Analysis: The Tale of Two Tiers
🦄 AI Unicorns Minted at Record Speed
🌱 Meet the Next AI Unicorns
💧 OpenAI’s $6B Liquidity Event
🥊 Perplexity’s $34.5B Chrome Gambit
🧠 Cognition's Blitz Continues with a $500M Raise
Let's jump right in.
AI funding went barbell this week.
At the top: two monster rounds.
Cohere and Cognition pulled in $500M each, together soaking up $1 billion, about 72% of all capital raised. The bet is clear: investors are doubling down on the “picks and shovels” of AI, foundational models, and developer platforms.
At the bottom: volume.
Across 40 total deals, $1.54B was raised, with dozens of smaller checks spread across verticals like AgTech, logistics, and enterprise automation. Specialised AI Niches and Enterprise AI Solutions were active, but the checks were modest, scaling proven business cases, not chasing moonshots.
Geography told the same story.
North America dominated: 25 deals, $1.3B, 84% of the week’s total. Cohere’s Canadian mega-round and Cognition’s U.S. raise cemented the region as the home of growth-stage AI capital.
Europe chipped in 7 early-stage rounds worth $72M, Asia logged 5 deals, and Australia added 2. The U.S. writes the big checks — the rest of the world seeds the pipeline.
Deal size brackets revealed the power law at work.
$100M+ (Mega-rounds): Just 2 deals (Cohere, Cognition) = 65% of total capital.
$50M–$100M: Zero activity.
$20M–$50M (Growth): 6 deals totalling $190M. Names like Squint ($40M), Profound ($35M), and Tahoe ($30M) are scaling fast.
$5M–$20M (Traction): The busiest tier with 16 deals—core Series A/B momentum.
<$5M (Early): 10 deals at seed and pre-seed. The pipeline stays healthy.
The takeaway: investors are making enormous bets on a few category leaders while still keeping the funnel of early innovation wide open.
The AI boom is minting unicorns at a pace that makes the dot-com era look pedestrian.
There are 498 AI unicorns now commanding a collective $2.7 trillion in paper wealth, according to CB Insights. The club added 100 new members since the start of 2023 alone..
2025 has already crowned 53 new unicorns, of which over half are AI-native. That’s faster than any prior tech wave, with startups now hitting $1B+ valuations in six years compared to seven for older cohorts.
OpenAI set the tone, with its valuation jumping from $300B to a reported $500B on the back of secondary sale talks. Anthropic's valuation is set to nearly triple to $170B in just five months.
This isn′t just creating valuable companies; it′s minting a new class of billionaires overnight. 29 AI entrepreneurs now hold a combined $71B net worth. CoreWeave’s co-founders are worth $1.7B each, Scale AI’s Alexandr Wang is worth $3.6B, while Figure AI’s Brett Adcock leads with $16.6B.
Still, there’s an even deeper story.
New AI unicorns are proving to be hyper-efficient, generating 83% more revenue per employee ($814k) than their older counterparts. That signals a fundamentally leaner, more scalable business model from day one.
The broader question for investors? How to distinguish sustainable growth from pure hype.
While the sheer number of unicorns seems frothy, the underlying efficiency metrics suggest a new paradigm of capital-light scaling. Startups like Clay are already generating over $1 million in revenue per employee, a figure that was once the exclusive domain of elite hedge funds, not tech startups.
All eyes now turn to the next wave of IPOs and M&A to see if these paper valuations can hold up in the public markets.
Is the next OpenAI hiding in plain sight?
Forbes’ annual Next Billion-Dollar Startups list is out, and AI dominates this year’s edition. Out of 25 companies spotlighted, 20 are building with artificial intelligence at their core, from military-grade autonomy to voice data pipelines.
Some early standouts:
Decart is the breakout star. Aiming to rival OpenAI, it was crowned a unicorn right after the list was published, raising $100 million at a $3.1 billion valuation.
Forterra is winning big in defence tech. The startup has already secured a $93 million U.S. Army contract to build AI-powered mine-clearing robots.
Gamma is tackling enterprise productivity. Its AI presentation tool has already seen 50 million global users and landed early customers like Amazon.
But not everything is a foundational model play.
The list highlights the shift toward vertical AI. Companies like AcuityMD (medical device marketing), Rogo (banking automation), and Basis (AI accounting) are applying AI to specific, lucrative niches.
And the revenue numbers are starting to build.
Defense startup Apex is already projecting $60 million in 2024 revenue.
AI presentation builder Gamma is on track for $16 million.
The new unicorn Decart is expected to hit $20 million.
Looking ahead, this list serves as a powerful leading indicator. While the tech giants battle for AI supremacy with nine-figure CapEx budgets, these startups are the nimble insurgents building the next layer of the AI economy. Their trajectory will signal where the next M&A and IPO hot spots will emerge.
OpenAI just announced one of the largest secondary sales in tech history.
The company is preparing a $6 billion secondary stock sale that values it at a staggering $500 billion, a massive jump from its $300 billion valuation in March. The deal will allow employees to cash in, providing a crucial tool to retain talent amid a fierce AI bidding war.
But the blockbuster deal arrives on the heels of a rocky product launch.
The rollout of its new flagship model, GPT-5, has been bumpy, with users complaining about losing access to features from prior models.
CEO Sam Altman admitted on X that the company "for sure underestimated how much some of the things that people like in GPT-4o matter to them."
Still, there is enormous investor optimism.
The secondary sale, backed by investors like SoftBank and Thrive Capital, provides huge liquidity to employees, a key advantage in retaining talent.
It follows an already massive $40 billion funding round led by SoftBank, of which $8.3 billion has already been secured.
OpenAI expects revenue to triple this year to $12.7 billion, up from $3.7 billion in 2024.
The broader concern? Balancing breakneck innovation with user satisfaction.
While OpenAI continues to push the technological frontier, the rollout of GPT-5 shows the challenge of managing a massive user base that has grown attached to its products. As the company scales to unprecedented valuations, its ability to execute smooth product transitions is now under the microscope.
All eyes now turn to see if the GPT-5 experience can be smoothed out to justify its half-a-trillion-dollar price tag.
Perplexity is swinging for the fences.
The AI search startup stunned Silicon Valley this week with an unsolicited $34.5 billion offer for Google’s Chrome browser, a price nearly double its latest valuation of $20B. The move is a direct challenge to the incumbent, timed perfectly as the DOJ pushes for Google to divest the asset on antitrust grounds.
But the bid is being met with widespread scepticism.
Many in the venture community have dismissed the offer as a brilliant marketing stunt rather than a serious M&A play. After all, the offer is nearly double Perplexity's own valuation, and Google hasn't signalled any intent to sell its crown jewel.
Still, Perplexity's underlying business is surging.
The company's annual recurring revenue has soared past $150 million, a more than 4x jump from its $35 million ARR just a year ago. Just last month, it launched its own AI-powered browser, Comet.
The broader concern? Whether this audacious bid can translate into market share.
While Perplexity has proven it can generate headlines, taking on Google requires more than just capital. The real fight is for the browser, the main gateway to the internet. This bid puts Google in a corner: ignore a well-funded challenger or legitimise them by engaging.
All eyes now turn to the Department of Justice to see if they will force Google's hand.
Cognition's dominance in the AI coding race keeps accelerating.
The company has now secured nearly $500 million in fresh capital, catapulting its valuation to $9.8 billion, more than double its $4 billion valuation from just months ago. The round, led by Founders Fund, solidifies its position as a clear frontrunner in the AI developer tool market.
That capital infusion is fueling serious momentum.
The financing follows Cognition's aggressive acquisition of Windsurf, a rival with a strong $82 million revenue run rate. Cognition swooped in after deals with OpenAI and Google fell through, absorbing Windsurf's IP and its remaining 200 employees to bolster its own AI agent, Devin.
The company's culture is just as intense as its growth:
CEO Scott Wu has mandated an 80+ hour, six-day workweek for the newly acquired team.
In a memo to staff, Wu declared, "We don't believe in work-life balance," framing the mission to build the future of software engineering as an all-consuming effort.
Why it matters:
AI coding is becoming one of the hottest battlegrounds in enterprise AI.
Cognition now controls both sides of the market: AI agents (Devin) and AI IDEs (Windsurf).
That dual positioning could make it the first true challenger to OpenAI, Anthropic, and Cursor.
But there are risks. Devin’s early reviews flagged accuracy issues, and Windsurf recently lost access to Anthropic’s Claude models (which Cognition says it will restore). Execution, not hype, will decide whether this $500M bet pays off.
Cognition is no longer just a flashy AI agent startup. With fresh capital and Windsurf under its belt, it’s positioning itself as the go-to platform for the future of AI coding, one where 20%+ of coding workflows could be automated by agents as early as 2026.
Got feedback, a story worth toasting, or a wild tech question? I'd love to hear from you; just reply.
