Headlines this week - Aug 10, 2025

A look at how capital is being deployed across future opportunities

C Gavilanes

Aug 10, 2025

This week in the future:

1 - The race for "superintelligence" is on, but the winner is far from certain

History suggests that today's AI leaders may not be the ultimate winners. A WSJ opinion piece draws a parallel to the 1990s browser wars, using Netscape's rise and fall as a cautionary tale. It argues that while today's AI innovators are market darlings, history shows that the initial front-runners in a tech revolution often don't win in the end.

Instead of one winner, we may see a proliferation of AIs. The future may be "polytheistic, not monotheistic," according to an essay by a16z’s Balaji Srinivasan. He argues that we are already seeing many competing models converge on similar capabilities, which suggests a future with a balance of power between various AIs rather than a single, dominant superintelligence.

Geopolitical blocs are likely to develop their own distinct AI ecosystems. E.g. China. Different geopolitical powers are already forging their own paths, with China being the most prominent example. As The Economist reports, China's AI ecosystem is now rapidly shifting its focus from pure model development to the widespread, practical application of the technology across its economy.

2 - Investor concern is growing over the sustainability of the AI infrastructure boom

Massive capital expenditure is straining Big Tech's finances. The sheer scale of the AI build-out is a "hidden risk to the economy," according to a WSJ article this week. The article notes that the massive capital spending required for infrastructure is now beginning to strain the vast cash pools of even the wealthiest tech companies.

Analysts are increasingly warning of an inevitable "bust”. A Bloomberg opinion piece this week argued that the current "AI building boom is bound to bust". While the "digital gold rush" is driving a windfall now, the author contends that hyperscalers cannot maintain this level of spending indefinitely, making a future downturn inevitable for investors to navigate.

3 - This is not yet affecting valuations: OpenAI's value soars to $500bn despite a mixed reception for GPT-5

A new share sale could value the company at $500bn, a valuation that could be justified by the unprecedented growth of ChatGPT. OpenAI is in talks for a new share sale that would value the company at $500bn, making it the world's most valuable private tech group. An FT analysis suggests this is justified by its unprecedented growth, as its flagship app, ChatGPT, is on track to reach 700m weekly active users, according to claims made by the company this week.

OpenAI has finally launched its long-awaited GPT-5 model. After a more than two-year wait, the company has unveiled its next-generation model, GPT-5. The Wall Street Journal reports that CEO Sam Altman has set high expectations, describing the new model as being like a "doctorate-level expert in any topic".

But the initial reception to the new model has been lukewarm. Despite the hype, the initial reaction to GPT-5 has been muted. The FT reports that many early users are calling the gains "modest" and the model "evolutionary rather than revolutionary," falling short of the hoped-for leap toward superintelligence.

4 - At the same time, consensus is growing on AI's potential to catalyze economic growth

AI is already creating massive hidden value for consumers. A WSJ opinion piece by two leading economists argues that AI's economic impact is being overlooked. The authors estimate that the technology is already providing nearly $100bn in "consumer surplus”, a value that, while not yet in GDP figures, reflects the fact that AI is already making people’s lives more productive.

If it materializes, this growth could actually be handy to solving the US debt crisis. AI-driven productivity may be the "secret sauce" for America's fiscal problems, according to a the FT’s Gillian Tett. Her article highlights an optimistic US budget forecast suggesting that even a modest productivity boost from AI could be enough to stabilize the national debt without austerity measures.

5 - This economic transformation will require profound changes in how we work

We need to reinvent critical thinking to “partner” with AI models. A new blog post at AI Salon this week argues that as AI automates more cognitive tasks, humans must adapt by developing a more "algorithmic mind". This new form of critical thinking will be essential for navigating a world where AI partners with us in our daily activities. In these emerging “partnerships”, humans would have to play a role as “prompters” (i.e. defining the problems to be attacked) and as “verifiers” (i.e. checking the outputs). For both things, new skills will be welcome.

This may require a radical overhaul of our educational systems. An FT op-ed suggests that to boost long-term growth, we should focus on "training the brain" from a young age. The article highlights research showing that even modest cognitive training for children (including the ways they interact with AI) can have an enormous impact on their future life outcomes.

Meanwhile, the negative impact on jobs has already begun. A "quiet havoc has begun" in the job market as a direct result of AI, according to a recent WSJ opinion piece. The author notes that this is no longer a future prediction, but a present reality that is already reshaping the world of work.

And the pace of AI-driven job cuts appears to be accelerating. An FT analysis this week warns that "the AI job cuts are accelerating". Major companies are now openly linking thousands of layoffs directly to AI, framing the reductions as a sign of progress and efficiency.

This disruption is now reaching high-paying, white-collar professions:
- Strategic Consulting: The Economist argues that strategy consulting is ripe for disruption, as AI models begin to produce the kind of frameworks that were once the exclusive domain of elite firms.
- Finance and Investment Banking: Some finance jobs would also be at risk, according to a new report in the FT. The article highlights how AI can already replicate a significant portion of a traditional financial analyst's job, or even substitute quants (something pretty unexpected not long ago…)

6 - Palantir has just confirmed its leadership in enterprise AI

Stellar second-quarter results surpassed $1bn in revenue. The company's sales surged, crossing the $1bn quarterly threshold for the first time. This "blistering pace" reflects a dramatic increase in both corporate and military investment in its AI tools, pushing the stock up significantly.

However, the company’s exuberant valuation is difficult to justify. An FT’s Lex Column this week argues that Palantir is actually trading at a nonsensical price. Even with the company’s impressive growth, the article notes its valuation multiple is in "exuberant" territory, far exceeding that of other successful tech giants.

Close ties to the Trump administration are seen as a key factor in Palantir’s success. The company's deep connections to the Trump administration might have been crucial to its growth, according to The WSJ. This relationship has helped it win over Washington and secure major government contracts, which in turn has fueled its massive 600% stock increase.\

7 - A battle is brewing in the consumer AI space between Apple and Meta

Meta and OpenAI are planning new AI-centric hardware. A WSJ article this week suggests that both Meta and OpenAI are developing new AI-native devices. This move is seen as a direct challenge to the iPhone's dominance, signaling a new hardware war centered on AI.

The company is now preparing its own push into generative AI. In response to this "existential risk”, Apple is ramping up its own AI efforts, with a new "Answers" team reportedly developing a ChatGPT-like product. Bloomberg reports that this new team is a sign that the company is taking the competitive threat from generative AI seriously.

In this context, Apple's focus on privacy could be a double-edged sword. As the AI race heats up, Apple's long-standing commitment to privacy reinforces the company’s singularity vs. rivals. While this stance resonates with consumers, it also could make it more difficult for the company to use the vast amounts of data generated by their users to power the most advanced AI models.

Also, the effort is happening as Apple is losing key AI talent This strategic shift is being hampered by a "string of departures" in the company's AI teams, as detailed in the FT. The article notes that rivals like Meta are successfully poaching top researchers, creating a crisis of confidence and hindering Apple's ability to keep pace.

8 - Longevity startups are building the future of healthcare, partially with the help of AI

The longevity sector is attracting intense investor interest (and generating controversy). Longevity startups are the subject of a major profile in The New Yorker this week, detailing the immense capital flowing into the space. The article explores the big ambitions and personalities within some industry startups like Altos Labs or Retro Biosciences, as well as the controversy and ethical questions that surround it.

Companies are working to create a "next-gen healthcare" model. Startups like Lifeforce are building a new model for proactive healthcare, as detailed in a recent Ashley Vance newsletter (& podcast). By using regular blood tests and health coaching to improve a user's "Lifescore," the company aims to move beyond traditional reactive medicine to a continuous optimization of health.

New innovation hubs for these technologies are emerging, A WSJ report highlights how longevity firms are pushing for Montana to become a new hub for biohacking and experimental treatments. The state's "right to try" laws are creating a favorable regulatory environment for companies to develop and test new, cutting-edge therapies.

9 - The US U-turn on AI chip sales to China keeps sparking intense debate

Nvidia has now officially received its license to export to China. Weeks after the policy U-turn, the US Commerce Department has started issuing licenses to Nvidia to export H20 chips to China. The FT reports the decision came just two days after Nvidia CEO Jensen Huang held another meeting with President Trump at the White House.

Analysts are still split on whether this helps or hurts US interests. The decision is highly controversial. A WSJ op-ed warned this week that the policy is "self-defeating" and will accelerate China's AI capabilities. The counterargument is that a continued ban would simply force China to develop its own domestic chip industry even faster, ultimately creating a stronger competitor.

Yes, China could always push for self-sufficiency, but they’re facing obstacles for that. Building a fully independent chip sector has proven difficult for Beijing. The FT reports that a state-led push to consolidate the fragmented domestic industry has hit roadblocks, with disagreements over valuation and ownership structure hindering the creation of "national champions".

10 - Robotaxis are already disrupting ride-hailing, with Waymo and Tesla as early winners

Robotaxis are rapidly gaining market share in early launch cities. In cities like San Francisco, robotaxis are quickly capturing a significant share of the taxi market, according to a new blog post by Tomás Pueyo this week. The analysis suggests that Waymo, and later Tesla, are positioned to be the biggest beneficiaries of this major disruption to urban transport.

Ride-sharing incumbents are preparing for an "existential battle”:
- Uber is now "readying itself for the driverless age”, reports The Economist. After previously abandoning its in-house efforts, the company is now partnering with players like Waymo, preparing for what its former CEO once described as an "existential threat" to its business model.
- Lyft is looking to international partnerships. To catch up with fast-growing rob taxi rivals, Lyft has decided to partner with Chinese tech giant Baidu, to launch robotaxi services in Europe, signaling that the competitive landscape for autonomous mobility is now global.