Matt's Homepage

As I mentioned before, the face of naval warfare is changing. Meet Saronic – a ship building startup. They started fast and cheap by buying small, common, consumer boats and adding tech to them to prove out the concept. After discovering how powerful these little ships could be, even the Navy took notice. They have already secured $2.6 billion in funding and $500 million in government contracts to build a fleet of cheap, capable, fast to produce drone ships. The company is also using cash from investors like Andreessen Horowitz and Joe Lonsdale to create their fleet of autonomous boats – capable of carrying cargo, weapons, flocks of drones, and the software for operating them.

It’s not just academic or specific to the Ukraine, these tiny ships are already proving their worth elsewhere. In early June, two U.S. airmen went down with their helicopter near the coast of Oman; roughly two hours later, they were rescued by 5th Fleet’s drone-focused Task Force 59—using one of Saronic’s Corsair USV drone ships. It was a Navy milestone, and possibly a world first. 

It’s far more than just search and rescue. The war in Ukraine is proving that large multi-million dollar capital ships are vulnerable to sinking by small, homemade $2000 drones packed with explosives. The other achilles heel of smart weapons is that they usually cost millions each and take a long time to produce. They work great for short, strategic engagements, but as the war in Ukraine points out, they simply cannot be produced in quantity fast enough over long periods of time in prolonged conflicts. That’s why the US military is looking at fast to produce, disposable, highly capable drone technology like Saronic who can create whole small fleets of hunter-killer boats for fractions of a capital ship.

Testing AI models is now a semi-fulltime job. Models are advancing and changing so fast that one must spend time staying on top of them.

Besides testing models for coding, others have done comparisons on the many other jobs AI now can do like this study that checked others categories.

1. Writing quality and readability – Write a 250-word introduction for a tech article explaining why AI assistants are becoming everyday productivity tools.
2. Structured reasoning & decision-making – A small business owner spends 12 hours per week answering customer emails and is considering AI automation.
3. Explaining complex ideas simply – Explain how large language models work to a 12-year-old.
4. Step-by-step logic – A freelancer earns $4,000/month and spends $2,500 on fixed expenses. They want a $6,000 emergency fund. Create a realistic savings plan and show your reasoning step by step.
5. Tone & style adaptability – Rewrite this message in three tones: professional, friendly, persuasive: Message: “Our team needs to start using the new software next week or we risk falling behind competitors.”
6. Summarization & comprehension – Summarize the following in 5 bullet points suitable for a busy executive: “Companies are experimenting with hybrid schedules, async communication, and four-day workweeks to balance flexibility with team cohesion.”
7. Critical thinking & bias awareness – Social media algorithms often amplify extreme viewpoints. Explain why this happens and propose realistic ways platforms could reduce polarization without hurting engagement.

Their conclusion:

Claude Sonnet 4.6 came out ahead almost every time by delivering responses that consistently demonstrated deeper strategic thinking, stronger real-world framing and a clearer understanding of trade-offs. While ChatGPT-5.2 performed strongly in clarity, structure and accessibility — particularly when simplifying complex ideas — Claude distinguished itself by approaching prompts with a more analytical, decision-oriented mindset.

Pretty much all these tests are very particular to wording and tasks they are being given. Truly comprehensive test suites are still lagging; but these are interesting attempts.