New Ask Claude or ChatGPT about any stock, using Bullpen. Bullpen in Claude or ChatGPT. 7 days free

Friday, November 20, 2026

Back to the front page

From X

Convequity on X

The test of the Jensen story is what Nvidia does about open weights next Jensen Huang resigned from $NVDA on 11 Nov after a spy story the Justice Department says it has reached no conclusion on. That news is old. The live question is whether his open-weight stance made him a target. The usual rebuttal is that Nvidia is buying about $10B of Anthropic ($ANT) stock, so Nvidia and the frontier labs are on the same side. We don't blindly accept that. The stake aligns Nvidia, the company, with Anthropic. It says nothing about Jensen, who was the most vocal champion of open-weight models and put Nvidia's balance sheet behind US open-model labs such as Reflection and Thinking Machines. ANT's own prospectus names open-weight models as a source of pricing pressure. A Nvidia that holds $10B of ANT and no longer has Jensen has a reason to stop funding that pressure. Look at who carries weight in Washington. Anthropic spent $5.6M on federal lobbying in 2026 through September. Nvidia spent $4.69M and OpenAI just under $3.8M. A lab outspent the company at the centre of chip policy. We think Nvidia's real influence in Washington was Jensen himself, and that has gone. No reporting ties any lab to his exit and we have no evidence of it. What Nvidia does next on open weights will show whether the concern is justified. Over the next few months we are watching three things: • Whether the open-model deals Nvidia was expected to join or lead still close. • What the permanent CEO says about open weights. • Whether Nvidia's pitch moves from enterprises running their own models to serving Anthropic and OpenAI. If Nvidia keeps backing open weights, the suspicion weakens. If it retreats, the suspicion hardens. A retreat would also be poor strategy. ANT's firm commitments for $GOOGL TPUs exceed $280B. Its Nvidia GPU contracts run up to about $200B, and the largest becomes cancellable on 90 days' notice from early 2027. Google sells Anthropic the chips, is buying about $10B of ANT stock itself and competes with Gemini. Nvidia's $10B is also small next to what its rivals hold: Google and Amazon together own about 28% of Anthropic, worth roughly $300B at the IPO price. Open weights spread Nvidia's demand across a broad base of enterprises that run their own models. Frontier labs have the most alternatives because they are big enough to design their own chips or split orders across suppliers: Google builds its TPUs in-house, OpenAI is designing accelerators with Broadcom, and Anthropic buys Google TPUs, Amazon Trainium and AMD GPUs alongside Nvidia. If Nvidia neglects its open-weight enterprise customers to lean on a few frontier labs, it ends up depending on customers such as Anthropic that are also buying Google's TPUs, and that helps Google.

Convequity on X

Gemini 4 Neon ranks first on coding benchmarks. After testing the Gemini 4 models, I would not replace Opus with them for engineering work Google ($GOOGL) launched Gemini 4 Neon on 18 Nov at $3 per million input tokens and $15 per million output tokens. It ranks first on the most-cited independent coding-agent index with a score of 58.6, and X is arguing whether it dethrones Claude. Neon is a real step up for Google. It is far better than Gemini 3.8 Flash, and for the previous few quarters Google users only had a Flash-tier model to work with. Its writing is very impressive, and so is its world knowledge. In my testing currently, Gemini 4 behaves more like GPT-6.2 Astra than like Opus in engineering work. I see the same two problems I see with Astra. It over-engineers, which means it builds more than the task asked for. And it follows the instructions or the agent guide to the letter without using its own judgment, so it can be single-minded in a way that does not look intelligent. Benchmarks do not capture this. Google's own launch blog shows how much a coding score depends on the setup. On the SWE-sweep test, the same Neon model scores 12.8% in a basic harness and 39.0% inside Antigravity, Google's own agent product. A benchmark score records whether the task was completed in that setup. It does not record whether the model built only what was needed, or whether it knew when to depart from its instructions. I call that taste, and we need a benchmark or some other way to verify it. I have not seen a dramatic take-up in Gemini usage or an oversubscribed Antigravity, and I think this is the reason. OpenAI and Google have the compute and the know-how to train a larger model and reach the top tier of the benchmarks. In daily use their models are still not comparable to Anthropic's. Anthropic is now public as $ANT, and I expect its competitive position to stay strong for the next two quarters.

Google

Google has narrowed OpenAI’s chatbot lead through Gemini distribution across Search, Android and Chrome, backing its AI stack with custom…

Key Man Risk

Meta and Nvidia show how founder-led strategy makes key-man risk central to long-term AI and Reality Labs spending.

GFL · long

Gfl Environmental INC

GFL Environmental, North America’s fourth-largest waste manager with a US$19 billion market cap, is pitched long.

The Week That Was

Treasury yields surged as the Fed’s expected hike shifted to December, while Micron beat guidance and forecast $61.5 billion in quarterly…

Nike, Part Two: The Deep Dive

Nike’s second Latticework program examines the company and the global athletic footwear and apparel industry for professional investors.

Publications we follow

Every publication on this site, hand-picked.

130 publications

Newsletters 110

Seeking Alpha 98 of 301

X accounts we follow

Every account on this site, hand-picked.

94 accounts

Investor letters we follow

383 firms

Sectors

Ordered by how often our editor reaches for them.