🤖🇨🇭AI & Machine Learning @ HSLU

Artificial Intelligence & Machine Learning at The Lucerne University of Applied Sciences and Arts (HSLU) in Switzerland.

[nxthompson @nxthompson]

This is a pretty striking shift toward Chinese models by American AI startups since the start of the year. https://t.co/80uaCWQCRE

[clem 🤗 @ClementDelangue]

Token costs are why there will be no saas apocalypse / good dev tools are cached intelligence for agents!

The popular theory goes: agents can write code, so they'll just rebuild every tool from scratch and hit raw APIs. no more dev tools, no more CLIs, no more software layers. just agents and endpoints!

We just tested this and the data says the opposite. We benchmarked Claude Code and Codex on real Hugging Face Hub tasks (~1,000 graded runs), with two setups: the agent-optimized hf CLI vs the agent hand-rolling curl or SDK calls from scratch.

Hand-rolling burns up to 6x more tokens on multi-step tasks and fails more often (84% vs 94% task success).

And that's just dropping one abstraction layer. It would obviously be orders of magnitude more tokens and a dramatically higher failure rate if the agent tried to bypass HF altogether and rebuild model hosting, versioning, and distribution from scratch. Every time an agent re-derives a workflow from raw API calls, you pay for that reasoning in tokens. every single run. a good CLI compresses that entire chain into a few high-level commands the agent can't get wrong.
In a world where everyone is complaining tokens are too expensive, abstraction is leverage: thousands of hours of design decisions your agent doesn't have to re-reason about at inference time.

Good tools are cached intelligence for agents!

So no, agents won't rebuild everything from scratch. they'll gravitate to the most token-efficient tools, because that's what their owners pay for. The software that survives won't just be accessible to agents, it will be accurate and cheap for them to drive.

We're seeing it happen with HF, which is becoming the platform for agents to use AI: ~49M requests in just two months, and growing fast!

https://t.co/Y7q6yuxZrZ

[Ethan Mollick @emollick]

It does seem like meaningfully better AI releases are accelerating, especially from OpenAI & Anthropic.

To illustrate, I caused this timeline to be created. It only lists new models that scored 3 points or higher over previous models in the Artificial Analysis index.

[kache @yacineMTB]

if you're doing AI research, do it on a single GPU with models that train in under 30 seconds

[Mark Carney @MarkJCarney]

We just launched Canada’s new AI Strategy: AI For All. 

We’re taking control of our future — with AI that’s governed by Canadian values, AI that’s accountable to Canadians, and AI that serves all Canadians.

[Thierry from arvy 🇨🇭 @ThierryBorgeat]

51% of the S&P 500's market cap is in stocks trading above 10x sales.

Half the index.

In 2002, after Sun Microsystems crashed 90%, CEO Scott McNealy famously said this about his own stock at 10x sales:

"At 10x revenues, to give you a 10-year payback, I have to pay you 100% of revenues for 10 straight years in dividends. Zero costs. Zero R&D. Zero taxes. Zero employees. What were you thinking?"

He was explaining why investors had been insane to pay it.

Today, half the S&P 500 trades there.

Different decade. Same math.

[Glenn Beck @glennbeck]

Saying "NO AI data centers" won't end in victory. They'll just build them anyways.

If you truly want to make a difference, you need to say, "No data centers, UNLESS..."

Turn it into a negotiation instead of a war, so we can fight over what really matters. No data centers UNLESS they're not draining our energy.

In fact, you could say, "you want 10 gigawatts of power? Ok, YOU pay for it. And because we’re allowing you to do it, you have to produce 15 gigawatts of power and give the excess to the grid."

It's a win-win and YOU control the terms.

[Anthropic @AnthropicAI]

The speedup isn’t just in volume. On open-ended coding problems where answers are unclear, Claude’s success rate is now 76%—a 50 point jump in just 6 months.

Many engineers also say Claude’s code quality is now on par with human code; we expect it to be better within the year.

[Hedgie @HedgieMarkets]

🦔UC Berkeley's computer science department just posted its worst failure rates in years. 35.3% of CS 10 students got F's in spring 2026, up from under 10% in prior semesters. Professor Dan Garcia says the primary driver is a "vast increase in academic dishonesty" through LLMs. Students use AI to complete assignments, never learn the material, then fail exams. His office hours, once full, are now empty.

My Take
Companies are firing experienced engineers while the pipeline that produces new ones is being gutted by the same technology. Students use AI to bypass the hard part of learning, show up to exams without the understanding, and fail. One professor discovered a student's linear algebra class had an "open AI" policy for homework and exams. That student then couldn't do basic linear algebra in the next course.

Both ends of the workforce are eroding at the same time. Senior engineers are getting cut to fund AI spending. Junior engineers are graduating without the skills because AI did their coursework. And the companies spending trillions on these tools haven't connected those two facts yet.

Hedgie🤗

[Mechanize @MechanizeWork]

We are now seeking a puzzle maker to help us create puzzles that LLMs can't yet solve.

[Will Kinney @WKCosmo]

ChatGPT is waaay better at physics, but good for @UChicago. Enterprise AI should be the norm at any research university.

[signüll @signulll]

the next massive consumer ai opportunity is making personal agents feel as intuitive as an iphone.

this is deeply important because this is the new software layer for everyday life.

most ppl do not want to configure workflows, manage prompts, route models, or think about agents at all. they want software that just works & the winning products will hide almost all of the complexity with taste incl. context, memory, & orchestration.

e.g. there’ll be baseline personal agents that come alive out of the box which are already understanding your context, patterns, relationships, preferences, apps, devices, routines, etc. then there’ll be ephemeral agents that spawn dynamically from intent, ambient capture, conversation, location, screenshots, email, calendar, camera roll, whatever. this is the software that assembles itself around the moment just like weather updates based on your location but way more in depth.

today even the most state of the art agent products feel like giving normal people shell access to a distributed system.

apple won by turning computers from something you operated into something you experienced. personal agents require the same transition.

whoever solves this becomes the ambient operating system for human life. small category btw.

[Oded Rechavi @OdedRechavi]

Most experiments fail, and negative results rarely get published. This means LLMs are unaware of the outcomes of most experiments.

[Matthew Prince 🌥 @eastdakota]

Welp, that happened faster than I predicted. Thought it would be end of 2027, then early 2027, but agentic traffic growing so fast that bots have now passed human traffic online for the first time in the Internet's history. https://t.co/2zX5bHdhsa

[Polymarket @Polymarket]

JUST IN: Scientists say AI has decoded communication patterns in mice, dolphins, apes, birds, whales, & cuttlefish — could eventually lead to humans communicating directly with animals.

[Ideogram @ideogram_ai]

Introducing Ideogram 4.0: the best open image model in the world.

Think it. Make it. Own it.

Download the weights, fine-tune on your own data, and run it on your hardware. Live on every Ideogram plan and the API today.

[Aoden Teo @AodenTeoMT]

Today, we’re excited to introduce Miso One, the most emotive voice model in the world.

Miso One is an 8-billion-parameter text-to-speech model for highly expressive speech generation. It emotes like a human and responds faster than a human, with just 110 milliseconds of latency.

We’ve open-sourced the model weights, with API access coming soon.

Hear how Miso One sounds in the thread below.

[Google @Google]

Today we’re introducing Gemma 4 12B — our latest open model that brings advanced agentic reasoning, vision and audio directly to your laptop.

It delivers performance nearing our larger Gemma models with a much smaller total memory footprint, while being small enough to run locally with just 16GB of VRAM. It’s open and accessible for everyone to use under a permissive Apache 2.0 license.

This is all made possible by our new, unified architecture that removes separate multimodal encoders. Here’s how we did it 🧵

[Ethan Mollick @emollick]

Law professors wrote questions they were asked during office hours. Gemini 2.5 & humans answered them then other law professors blindly judged the results:
-Gemini had a 75% win rate vs. professors
-Gemini's answers were rated LESS harmful than humans
-Newer models do even better

[terminally onλine εngineer @tekbog]

mythos find all the laws that google, meta and openai break so we can fine them, make no mistakes