Few Stars, Lots of Money

Holaaaa, Oláaaaa, Helloooo!

Friday, Halloween eve, and what spooked us this week wasn't a costume: it was the zeros.

An AI lab barely over a year old raised $1B with no product on the market. A megafund closed $10B to bet on about ten companies. And DeepSeek, according to Reuters, raised something close to $15B. Meanwhile, in Brazil, the question of the moment is who gets into the funnel. Grab your coffee (or your pumpkin) and let's get to what matters: the Deal Brief, jobs, events and a meme to ease the scare. Let's go, captain!

But first, subscribe to the newsletter for free to receive it every Friday:

And if you want to follow the daily news on WhatsApp, follow the channel!

Fact: Anthropic filed a Form S-1 with the SEC for a public offering of Class A shares, intending to list on the Nasdaq under the ticker ANT. The number of shares and the price range have not yet been set. A confidential draft had been submitted in June. In H1 2026, revenue was $16.3B, versus $1.2B in the same period of 2025. As of September, more than 11,900 customers were spending $100K or more per year. The offering price and size have not yet been disclosed. According to Yahoo Finance, the net loss was $80.7B through June, largely due to accounting charges. NPR cites reports of a valuation between $1.5T and $2T, a third-party estimate.

Analysis: this is the first time a developer of advanced AI models has opened its numbers to the market, offering a rare look at the economics behind the technology: revenue concentrated in large enterprise customers and the governance challenges of a fast-growing sector.

Sources: Anthropic, Oct 26 · CNBC, Oct 26 · NPR, Oct 26

Fact: Reuters reported that DeepSeek raised about 100 billion yuan ($14.98B), with Meituan, Geely and GigaDevice among the investors. The data comes from anonymous sources, and the company had not publicly confirmed it in the excerpts we read.

Analysis: The capital comes from industrial companies and platforms, not classic VCs. The competition with American labs is now also a competition for strategic capital.

Source: Reuters via MarketScreener, Oct 22

Fact: A $237M Series E at about a $3B post-money valuation, led by Bristol Myers Squibb. It came together with a collaboration with BMS and Tempus AI on neurological diseases. Pathos will license Tempus data for $55M over five years. BMS gets non-exclusive access to the platform and three years of exclusivity in neuroscience. Clinical results are not yet proven; for now, this is an announcement.

Analysis: The pharma company leads the round and brings the platform into its own R&D. Proprietary data becomes the core asset.

Sources: Press release (BioSpace), Oct 22 · Fierce Biotech, Oct 23

Fact: A Series B led by GM Ventures to scale US manufacturing and enter the data center market. The 4 GWh factory in California is expected to start production in Q1 2027. In the meantime, the company uses cells from Chinese suppliers, and sodium-ion production remains concentrated in China. The backlog of about $1B in orders is reported by the company.

Analysis: Storage for AI's energy demand is a central theme of the year. Dependence on the Chinese supply chain is the critical point.

Sources: pv magazine, Oct 23 · Power Technology, Oct 22

Fact: According to the Q3 2026 analysis, the number of equity rounds for Brazilian startups fell from 161 (Q1 2024) to 60 (Q3 2026). Decade raised $85M in a round classified as seed.

According to Sling Hub's Q3 2026 report (October 2026 edition) and Danilo Picucci's reading of it (O Elo, Oct 27, 2026), Brazilian startups raised $3.25B in Jan–Sep 2026 versus $3.09B in the same period of 2024. The number of equity rounds fell from 436 to 202. The average ticket rose from $3M (Q3 2024) to $5.4M (Q3 2026), or $4.1M excluding one $85M round. The total volume includes debt and FIDCs (receivables investment funds), which made up almost half of the volume in Q3. In Q3 2026 alone: 66 rounds and $629M (-63% on the quarter, -50% on the year); equity of $325M across 60 rounds; rounds with corporate participation totaled $202M (32% of volume, versus 4% in Q2).

Analysis: Capital is back, but concentrated in fewer companies. For founders, the bottleneck is getting into the group that receives a check; for investors, fewer deals with larger tickets.

Source: Substack (secondary), ~Oct 27

Fact: The report Barriers and Opportunities for Automation in UK Ports concludes that the country has not yet moved from pilots to large-scale commercial deployments, unlike parts of Asia and northern Europe. It points to mixed automation (autonomous and manual operations coexisting) as an opportunity.

Analysis: The bottleneck from pilot to scale is the same one faced by terminals and suppliers in Brazil.

Sources: AMTI · Computer Weekly (exact date not confirmed, between Oct 21 and 27)

Anyone who follows soccer knows that building a championship team isn't simply about signing the most expensive players. You can have the best strikers and million-dollar salaries, but without a solid foundation and a game strategy, you'll hardly win a title.

And what does this have to do with Venture Capital? More than it seems.

In recent weeks, we've watched a true billion-dollar transfer window in the artificial intelligence market.

A Dallas-based firm announced a $10 billion fund to invest in approximately ten companies. Flapping Airplanes raised $1 billion despite having no commercial product. And DeepSeek, according to Reuters, reportedly raised about $15 billion.

These are impressive numbers. But the question remains: are we building an innovation ecosystem or just funding a squad of superstars?

While billions are concentrated in a few AI labs, thousands of startups continue to struggle to access capital, validate technologies and win customers.

In Brazil, the picture deserves attention. Preliminary data points to a significant drop in the number of equity rounds, even with relevant volumes of capital invested. It's like celebrating a record transfer window while the smaller clubs stop developing new players.

Throughout my career working with startups, corporations and investors, I have seen promising businesses struggle not because they lacked good solutions, but because they lacked access to capital, markets and strategic connections.

And here is the point: innovation doesn't happen only where the money is. But money can determine which innovations get the chance to happen.

Not every startup needs to become a unicorn to generate economic impact. Solutions that increase productivity, reduce waste or transform traditional sectors also deserve a place in this game.

Perhaps the ecosystem's biggest challenge is not to discover the next unicorn, but to create conditions for more companies to compete.

After all, a championship isn't won with stars alone. You need a team.

And here's the provocation: are we investing in the next generation of innovative companies, or just competing for the stars the market has already chosen?

* Outside the 7-day window by 2 days; kept for its Brazilian relevance. Alice's efficiency and growth figures are reported by the company.

The Brazilian industrial AI startup announced on Oct 23 a $130 million Series D led by Sapphire Ventures and Catalyst (M&G's growth equity arm), with participation from N47, General Catalyst, Bosch Ventures, 53 Stations, Monashees, Endeavor Catalyst, Norte Ventures and Y Combinator. Annualized revenue is close to $100 million, about half of it in the US, and the technology runs in more than 2,000 facilities. The funds will go to physical AI R&D, LLM training and the purchase of Nvidia processors. Revenue figures are reported by the company.

Analysis: this is a case of applied AI selling to global industry from Brazil. For B2B founders, it works as a benchmark for metrics (ARR and revenue mix).

Sources: LatamList, Oct 23, 2026 and Techloy, Oct 23, 2026

According to Sling Hub's report (October 2026 edition), the third quarter had 66 rounds totaling $629 million, down 63% from Q2 and 50% from Q3 2025. The five largest rounds of the quarter:

Adding up the first three quarters of the report, equity rounds fell from 436 in 2024 to 202 in 2026, and total volume, which includes debt and FIDCs, was $3.25B versus $3.09B. Debt and FIDCs accounted for almost half of Q3 volume. In other words, "raising more" holds for the total, not for equity.

Analysis: capital is more concentrated in a few large deals and in structured credit. Corporations appear in 32% of the quarter's volume (4% in Q2).

Source: Sling Hub, Open Innovation & CVC Ecosystem in Brazil, Q3 2026. Further reading: O Elo, Oct 27, 2026

Events from Oct 29 to Nov 5: none verified.

Attribution according to Quote Investigator: Alan Kay claims to have coined the saying and says he has used it since 1971. Competing attributions circulate, but Kay's has a statement from the author himself.

Humor, not data or investment advice.