Most AI announcements are about the model itself: a new release, a better benchmark score, a lower price. Anthropic's latest news is different. It is about what the model found in biological data, and about who did the physical lab work that followed.
Only a week after confirming that it runs a wet biology lab in the Bay Area, Anthropic said the lab had already produced what the company considers a significant result. It describes the find as a new enzyme "system" with "properties reminiscent of CRISPR."
What was found: an enzyme system hidden in viral DNA
CRISPR began as a natural immune system in bacteria, which use it to defend themselves against viruses. Researchers later turned it into a gene-editing technology, and it is now widely used in labs.
Most of the debate about AI regulation in the United States has focused on disclosure rules, liability questions and voluntary commitments from developers. A new bill from Bernie Sanders and Greg Casar goes considerably further. It would stop the most compute-intensive AI development for a period and permanently outlaw one category of technology.
Sanders and Casar, a Democratic member of the House, presented the proposal on 23 September under the name Ban Artificial Superintelligence Act. It rests on three measures: a permanent prohibition on artificial superintelligence, a pause on the largest AI models, and a new federal department dedicated to AI.
Progress in AI agents has mostly been measured in capability: better reasoning, more fluid conversation and access to more tools. That is good news for a company that wants to automate customer service or place an agent inside internal workflows. It is less comforting for the security team that has to work out what the agent will do when someone tries to trick it.
A recent special edition of the AI Weekly newsletter looked at that tension. It paired a sponsored perspective from testing company Spec27 with six pieces of research on agent security. The common thread is that a smarter model does not automatically make a safer agent. As agents become more flexible and more deeply connected to data and tools, their attack surface grows with them.
Most of the talk about shrinking language models centers on 4-bit and 8-bit quantization. Ternarisation goes much further. It cuts every weight in a model down to one of three values: -1, 0, or +1.
That idea has been circulating in AI research for a while, most visibly through Microsoft Research's BitNet b1.58 work. It gets especially interesting at the scale of a model with around 1.85 billion parameters. A model that size is small enough to experiment with on modest hardware, but large enough to be useful for real tasks.
What ternarisation means
In a standard model, each weight is stored as a 16-bit or 32-bit floating-point number. Ternarisation replaces each weight with -1, 0, or +1, usually paired with a scaling factor that is shared across a group of weights or a whole layer.
When a chatbot types out its answer word by word, the speed you see is often limited by how fast a chip can move data, not by how fast it can calculate. That is why memory bandwidth has become one of the most important specifications in AI hardware.
Every token needs the whole model
Language models generate text one token at a time. For each new token, the chip has to read the model's weights, often billions of numbers, from memory into its compute units. For a single user, the math performed on each weight is small. As a result, the processor spends much of its time waiting for data to arrive.
CyrioX is financed by advertising. You can choose how you want to use this website:
With advertising: we load an advertising script from a third-party ad network. The ad network may set cookies, use your IP address and device information, and may process data outside the EU.
Ad-free for €0.99 per month: no advertising and no advertising tracking. Cancel at any time.
You can change your decision at any time via "Cookie Settings" at the bottom of every page.
