“Most mornings I learn more from New Items than I do from all of the traditional papers I read combined.” — Michael Blair, former presiding partner, Debevoise & Plimpton.
1. An experimental mRNA vaccine allowed people with high-risk melanoma to live longer, cancer-free, the pharmaceutical companies Moderna and Merck announced Wednesday. For decades, cancer vaccines have been one of the most tantalizing ideas in cancer treatment — ** offering the hope of harnessing a person’s own immune system to attack tumors**. “It’s a big deal,” said Dr. Elias Sayour, a pediatric oncologist at the University of Florida, who was not involved in the trial. The results, he said, “could help usher a whole new wave of new therapeutic treatments for cancer.” The mRNA technology that was instrumental in the rapid vaccine development in the Covid-19 pandemic added a new element of speed and flexibility to the quest for therapeutic cancer vaccines. Scientists can use the mRNA platform to create bespoke, individualized vaccines matched to the genetics of patients’ tumors. (
2. GeneralistAI:
Humans have a remarkable ability to perform new physical skills from only one or a few examples.: it can learn a new task in seconds, from a single example, without gradient updates or fine-tuning. It displays broad capabilities across one-shot and few-shot learning from demonstration, as well as zero-shot physical generalization. Although the tasks are simple and short-horizon, this is the first model we know for which one-shot and few-shot learning of physical skills have emerged at scale. We view these results as a significant step towards our mission of building general intelligence for the physical world. ([Our latest robot foundation model, GEN-1.5, exhibits the beginnings of that same ability]Source: generalistai.com. Wired article about this is.)[here]
3. Joachim Klement:
In my regular research (behind a paywall), I have been saying for a while that I think the future of AI is not large language models (LLM),but small language models (SLM) run on local desktop computers or even mobile phones. In May,that compared these SLMs with the performance of LLMs run in data centers. If their results are true, then[a team from Stanford University published research]. Seriously, if you are an investor trying to figure out where to invest in the AI hype, you need to read[we will hardly need any data centers in the future, and the hyperscalers are wasting hundreds of billions of dollars in investments][this paper]in full. (Sources: panmacmillan.com, klementoninvesting.substack.com, arxiv.org)
4. ** Hu Hong**:
Leading Chinese AI startups are facing immense pressure from impatient capital to demonstrate a path to profitability amid staggering cash burns. The brutal reality of the open-source business is that you are subsidizing not just your developers and users but your direct competitors.In an industry where open source and monopoly profits are mutually exclusive, no rational new entrant will invest billions in foundational models when it can simply free-ride on existing open weights. Capital markets are realizing this.
