Hugo Pinheiro

18.3K posts

Hugo Pinheiro banner
Hugo Pinheiro

Hugo Pinheiro

@user_ops

Building architecture patterns and agent skills to enable the coming personal agent microcloud, focusing on edge compute, thoughts and opinions are my own.

London, Ontario Katılım Eylül 2015
2.8K Takip Edilen742 Takipçiler
Sabitlenmiş Tweet
Hugo Pinheiro
Hugo Pinheiro@user_ops·
I'm looking for a new remote Devops, Platform engineer role, feel free to take a look at my resume - github.com/Opswatch/Resum… and personal projects works.ops.userops.me if I seem like a good fit, please dm me or reach out by email, cheers.
English
0
1
0
305
Hugo Pinheiro retweetledi
Matt Johansen
Matt Johansen@mattjay·
Woah. Cursor 0day dropping due to 200 days of them ignoring the disclosure.
Matt Johansen tweet media
English
34
71
2.1K
247.9K
Hugo Pinheiro
Hugo Pinheiro@user_ops·
@wavefnx @0xIlyy Pretty sure they originally did it because they thought rage bait would drive more traffic, turns out most of us just want to have good discussions on our preferred subjects 🤷‍♂️
English
0
0
2
6
wavefnx
wavefnx@wavefnx·
@0xIlyy >By 2026 we will have flying cars >slop-blocking extensions cause a social media site isn’t social Also was probably on purpose
English
1
0
7
98
Hugo Pinheiro retweetledi
Alexey Fateev
Alexey Fateev@superalesha·
Hy3, 295B params. 1-bit. 4x3090. it actually runs. then i found out that the default way to split the model across gpus was silently killing my decode speed on long context. same model, same 4 gpus, 32k context: layer-split (default): 16 tok/s decode, 375 prefill tensor-split: 27 tok/s decode, 317 prefill literally one flag. +72% speedup decode bravo @TencentHunyuan 🎉
Alexey Fateev tweet media
English
12
3
57
6.5K
Hugo Pinheiro
Hugo Pinheiro@user_ops·
@EdSander At least they tried I guess, I find China's retail operations fascinating, north America seems so boring from a retail perspective compared to what Chinese companies are trying.
English
1
0
2
14
Ed Sander | China Digital Retail Report
Premier probably came to late. They started preparing for it in 2021. But by 2023 people had already become more economical in their spending. Besides, with one store, Premier wasn't really worth keeping. Hema claims some if the learnings from Premier have been incorporated into Hema Fresh. I actually also see a lot of Hema Premier's approach in Meituan's Xiao Xiang (Little Elephant) offline supermarkets.
English
1
0
1
27
Ed Sander | China Digital Retail Report
ANOTHER ONE OF ALIBABA’S HEMA FORMATS QUIETLY DISAPPEARS Whenever I return to certain Chinese cities, I have ‘anchor points’ that show me how things are developing and tell me stories you won’t read in the news. Shanghai’s Longzimeng mall is one of those places. I have been visiting it for many years, and I have seen the ALDI format change, the Pet & Fresh pet food store by Alibaba’s Hema founder Hou Yi close almost as fast as it opened, and I have closely monitored the changes in Hema Premier. Today, during a 2-day visit to Shanghai, I dropped by the Longzimeng store to check on the status of Hema Premier. To learn what I found, read the full post on China Digital Retail Report. substack.com/profile/448473…
Ed Sander | China Digital Retail Report tweet mediaEd Sander | China Digital Retail Report tweet media
English
2
2
5
712
Hugo Pinheiro
Hugo Pinheiro@user_ops·
@heyandras @orcdev It has massively, mostly people I follow now, thankfully no more rage bait crap or politics.
English
0
0
0
13
OrcDev
OrcDev@orcdev·
I love this new algo
English
7
0
14
1.4K
Hugo Pinheiro
Hugo Pinheiro@user_ops·
@heyandras The good old days when one of my buddies dad worked as a water slide architect and had a Pentium for 3d modeling and we all huddled around it going 😲, all I had at the time was a old amd 486.
English
0
0
1
16
Andras Bacsai
Andras Bacsai@heyandras·
yes, I am old enough that I remember this was pure magic in that time my kids won't experience this or maybe they will, but it will be something different like today I talked with my 9 years old son on how we should make games with AI, he talks to AI about his ideas and let the magic happen
Ramin Nasibov@RaminNasibov

are you this old?

English
8
1
24
2.7K
Hugo Pinheiro
Hugo Pinheiro@user_ops·
@Dan_Jeffries1 Gemma models are actually good at it with some guardrails, but you still need guardrails ( think a harness but for writing ) or else its slop, most other models these days are optimized for agents and code, even Claude which used to be good for writing.
English
0
0
0
18
Daniel Jeffries
Daniel Jeffries@Dan_Jeffries1·
LLMs have most certainly gotten better at writing well and I love them but as a thirty years veteran writer I can still spot the patterns. They have a remarkably consistent structure and sentence style (block paragraphs, this and then this and this sentence patterns, lack of contractions, overly formal word choice, grammar that's too perfect instead of a style that knowingly breaks the rules to hammer home a point, etc). The other challenges are subtler. There is often no unifying idea or, worse, stray ideas find their way in that don't make sense in the overall structure and don't make the overall points cleanly or that don't flow smoothly from the last idea. To me every X post reads the same now. Same overly long read, puffy middle and semi-conversational "here's the thing" style. No doubt it's gotten better and if you know how to use it (like anything in life) it's an amazing tool but I find that when I start with LLM drafts I have to work harder sometimes, almost like being an editor of someone else's writing (someone who is slightly and subtly nuts) instead of it being my own words end to end and where the whole shape of the article or story and what it's doing is in my mind. But this is the worst it will ever be. Still some of the problems will persist. Ultimately if you don't do the work or part of the work to shape the LLM's writing (or code or art or whatever) then it's not yours and you don't understand it and can't command the language with the same thundering power.
English
19
9
87
7.3K
Hugo Pinheiro
Hugo Pinheiro@user_ops·
@just_cameron @tenobrus Honestly I don't know if Gemini is even at the sonnet level for regular stuff, it's good at frontend and frontend code but that's kind of it 🤷‍♂️🫣🫠
English
1
0
1
18
Tenobrus
Tenobrus@tenobrus·
google is sort of like anthropic if every anthropic release was sonnet 5
English
27
6
818
34.8K
Hugo Pinheiro
Hugo Pinheiro@user_ops·
@GergelyOrosz Gemma ones aren't horrible for writing with some rails, everything else pretty much sucks 😔. ( Haven't tried Gemini for story writing )
English
0
0
0
123
Gergely Orosz
Gergely Orosz@GergelyOrosz·
With every new model, I put it to the test in writing - to see if I could do what I do - I give it a bunch of interviews I did on a topic, and ask to write an article in my style. The models fail spectacularly, Fable and GPT-6 no different. Lots of words, but no understanding
English
46
7
315
35.3K
Maxime Rivest 🧙‍♂️🦙🐧
This seems to be about as nice the remarkable paper pro can animate it's screen. It still is weirdly mesmerizing to me. Inspired by Wenting's work with modos and his demoed using that bad apple clip.
English
9
2
74
6.7K
Hugo Pinheiro retweetledi
Dee
Dee@dee_hw·
1/ What if Frank Lloyd Wright designed the Personal AI Computer? It would look less like a PC and more like architecture. A tectonic pinwheel: four GPU masses in C₄ rotational symmetry around a central void, set on a service plinth and wrapped in a perforated envelope.
Dee tweet media
English
6
5
53
4.6K
Hugo Pinheiro
Hugo Pinheiro@user_ops·
GIF
Lenny Rachitsky@lennysan

My biggest takeaways from @noamseg, diving into the results of our large-scale survey of tech workers in 2026: 1. The #1 fear in tech right now is not losing your job to AI—it’s being squeezed to do more work for the same pay. When asked to identify their biggest concerns with AI, “losing my job to AI” ranked near the bottom of the survey results. What people are actually worried about is being squeezed: having to do more for the same pay, and watching the quality of their work slip. AI raised the bar for output, and the reward was...more work for the same pay. 2. Manager effectiveness is the biggest lever for employee well-being. Workers with highly effective managers report roughly 65% higher job enjoyment and dramatically lower burnout. The problem is that only 25% of respondents rated their manager as highly effective (and 36% rated their manager as ineffective). This finding has held consistent across both years of the survey. The squeeze employees are feeling is something managers are best positioned to protect against. 3. Most people in tech wouldn't recommend their own role to someone entering the industry today. Using an NPS-style question (scale of 0–10), not a single function scored above zero when asked if they’d recommend their career path to others. Founders came closest to neutral; designers and researchers scored worst. 4. AI is splitting the tech workforce in half. When asked how AI has shifted their professional identity, 50% of respondents said they feel amplified—more capable, more productive, more excited about their future. The other half feel their role is being redefined (27%), that they’re feeling destabilized (14%), or that they’ve been diminished (5%). Which category you’re in correlates with your career optimism, burnout, and layoff worry—an effect about three times as large as the next-biggest factor (manager effectiveness). 5. Burnout surged 10 points in a single year, while optimism fell 6 points. Shipping faster is burning people out. More prototypes, more PRDs, more agents, more output. As the report puts it, “the speed AI unlocked got plowed straight back into expectations.” The one glimmer: job enjoyment held roughly steady year-over-year. Many people are burned out and still having fun. As @nikhyl put it, many people are in a state of "smiling exhaustion." 6. Quality of people's work is suffering. 97.2% of respondents said AI is making them better at their job; close to 50% said “very much” or “extremely” better. But when asked what “better” actually means, the answer consistently came back as “more, faster”—not higher-quality work. Even more concerning: people reported a phenomenon Noam calls “cognitive rot.” They see the AI’s initial output, accept it without applying their judgment, and gradually let their own critical thinking atrophy. 7. Designers and researchers are the most negative group in tech, for the second year in a row. They lead in feeling destabilized or diminished by AI, reporting high rates of anxiety and overwhelm, worries about losing their roles to AI, and unwillingness to recommend their careers to others. Noam’s read: this doesn’t mean these roles are becoming irrelevant—if anything, he argues the opposite. Taste, craft, judgment, and the ability to create genuinely novel experiences remain stubbornly human. “The industry needs us,” he says. “This is a call to get in there and do our thing.” 8. Founders are still the happiest people in tech. For the second consecutive year, founders score highest on optimism, job enjoyment, AI excitement, and lowest on burnout and layoff worry. The “there’s never been a better time to build” narrative holds emotionally, at least for now. Founders have agency, autonomy, and direct relationships to the output. 9. Company size is linearly correlated with employee misery. Across burnout, optimism, layoff worry, and career recommendation, outcomes degrade in a straight line from 1-to-10-person startups to 5,000-to-10,000+-person enterprises. There’s no size at which things get better before they get worse again.

ZXX
0
0
1
70
Hugo Pinheiro
Hugo Pinheiro@user_ops·
@feross Your role is to now be the imagineer ( borrowed it from Disneys imagineering dept )
English
0
0
2
49
Hugo Pinheiro
Hugo Pinheiro@user_ops·
@Defence_blog Interesting that both it and the flyer are analogs to trucks and light vehicles in the dakar race 🤔😃
English
0
0
0
55
Maxime Rivest 🧙‍♂️🦙🐧
@user_ops one company! i didnt know that.. yeah i want to look into boox a bit. do you have an existing workflow that you like to work with ai through einks?
English
2
0
0
61
Hugo Pinheiro
Hugo Pinheiro@user_ops·
@MaximeRivest I been thinking about it I would probably say just to display things, let's say you have a bone conducting headset that you use to talk and control your agent, sometimes you need it to display things, that's where a small eink or phone like device would come into play.
English
0
0
0
27