Amicus AI

privately held Silicon Valley firm with a 5 year track record. No Bias Securities Research Reports

[DeepSeek @deepseek_ai]

🏆 World-Leading Reasoning

🔹 V3.2: Balanced inference vs. length. Your daily driver at GPT-5 level performance.
🔹 V3.2-Speciale: Maxed-out reasoning capabilities. Rivals Gemini-3.0-Pro.
🥇 Gold-Medal Performance: V3.2-Speciale attains gold-level results in IMO, CMO, ICPC World Finals & IOI 2025.

📝 Note: V3.2-Speciale dominates complex tasks but requires higher token usage. Currently API-only (no tool-use) to support community evaluation & research.

2/n

[DeepSeek @deepseek_ai]

🚀 Launching DeepSeek-V3.2 & DeepSeek-V3.2-Speciale — Reasoning-first models built for agents!

🔹 DeepSeek-V3.2: Official successor to V3.2-Exp. Now live on App, Web & API.
🔹 DeepSeek-V3.2-Speciale: Pushing the boundaries of reasoning capabilities. API-only for now.

📄 Tech report: https://t.co/7EyydyNuG0

1/n

[Amicus AI @AmicusAI]

Shake Shack and Stocks at SFSU!

[Amicus AI @AmicusAI]

The Mighty Ducks of AI: Scrappy Innovation
https://t.co/68iEKlpurL
How do you turn a bunch of "nobodies" into the Mighty Ducks?

We go through history, from Apple, to Facebook, to WhatsApp to OnlyFans. Scrappy teams win.

@deepseek_ai @Apple @Meta @WhatsApp @OnlyFans

[Amicus AI @AmicusAI]

The Mighty Ducks of AI: Scrappy Innovation
https://t.co/68iEKlpurL
How do you turn a bunch of "nobodies" into the Mighty Ducks?

We go through history, from Apple, to Facebook, to WhatsApp to OnlyFans. Scrappy teams win.

@deepseek_ai @Apple @Meta @WhatsApp @OnlyFans

[👋 Jan @jandotai]

DeepSeek R1.1 just matched Claude Opus on Aider's polyglot benchmark - 70.7% Pass@2.

Old R1 scored 56.9%, so this is a +13.8pt jump. Same test, same setup, posted by a user on r/LocalLLaMA.

Cost to run: ~$3 off-peak.

[Mcqueen42 @McQueenFu]

DeepSeek-R1-0528 is new sota opensourced llm. Let's support opensource.

[Leonidas Tam @LeoKTam]

Pleased to announce that our BrainDate session: AI in Finance, is fully subscribed at @nvidia NVIDIA's GTC: GPU Technology Conference 2025, taking place March 17th onwards. See everyone there. @NVIDIAAI #GTC25

[Amicus AI @AmicusAI]

1/25 $BWB A new Bank from Bridgewater that services mortgages in the Twin Cities. Return on equity is low at 7% and equity tier is only at 9% and trying to increase. Grade D. Confidence 0.6. https://t.co/xamuoxN3xe

[Amicus AI @AmicusAI]

1/25 $TMUS 7% earnings miss and 4% revenue beat. Should be about one percent decline in revenue. However, they’re saying they added over 900,000 households or lines and over 2 million for the year. They say they raised revenue per user by one percent. Grade c. Confidence 0.6.

[Amicus AI @AmicusAI]

1/25 $KLAC 5% earnings beat, and one percent revenue beat. Significant decline from last year’s earnings probably around 8% decline. And they are guiding flat for the next year. Grade F. Confidence 0.6. https://t.co/xamuoxN3xe

[Amicus AI @AmicusAI]

1/25 $V visa. 3% earnings beat and one percent revenue beat. 7% growth with 5% growth in the US and 13% growth in Europe. Some sort of partnership with Service now for dispute resolution. Active and almost every country and increasing merchant penetration by 17%. Grade B.

[Amicus AI @AmicusAI]

1/25 $INTC 20% earnings beat and 2% revenue beat. 10% growth and 230 basis points of gross margin expansion. They are quoting mobileye revenue however it should be independent. They mention EUV machines which like Kla Tencor might not be a promising foundry investment.

[Amicus AI @AmicusAI]

1/25 $VLY 2% earnings beat and 4% revenue miss. They somehow lost 2.1 billion in direct deposits and tried to rebuild it with 1.6 billion in inbound new households which they added over 10,000 households but primarily through promotional activity. Grade D. Confidence 0.6.

[Amicus AI @AmicusAI]

1/25 $COF 16% earnings miss and in-line revenue. One percent growth and increase in charge of 230 bases points while they held their net interest income steady. They are increasing reserves by another 2 billion to something ridiculous like 120,000,000,000.

[Amicus AI @AmicusAI]

1/26 $GNTX 4% earnings beat and 2% revenue beat. 19% growth on top of 350 basis points of gross margin expansion which they said 100 basis points were due to one time price increases.

[Amicus AI @AmicusAI]

1/26 $CL 3% earnings beat and one percent revenue beat. Growth around 8% and they have been increasing pricing aggressively for 7 quarters and they say that protein inputs are moderating so they’re pricing might start to moderate.

[Amicus AI @AmicusAI]

1/26 $NSC One Percent earnings miss and in-line revenue. 5% revenue decline with 3% increase in operating cost. 5% growth and intermodal volume is hiding their weakness in rail and possibly weakness in American industrial is driving railroad activity down. Grade D.

[Amicus AI @AmicusAI]

1/26 $FCNCA Possibly 5% earnings missing and they had a $67 million charge from the FTC assessment. Their return on equity is 13.8% and they experienced about 38 basis Points of interest margin compression. Grade C. Confidence 0.6.

[Amicus AI @AmicusAI]

1/26 $BAH 21% earnings beat and one percent revenue beat. 12.7% growth with 2% decline in the year. Backlog group by 15% and headcount group by 8%. Defense was particularly strong at 17% growth. Grade B. Confidence 0.6. https://t.co/xamuoxN3xe