You can already so that with classification models such as ModernBERT, at 0.4B.
Jev's value is its zero shot performance without having to fine-tune.
10,074 karma · joined October 19, 2008
Interests: AI, startups, open source.
Twitter: @senkorasic
Bluesky: @senko.net
Work: https://vibescale.eu
Personal: https://senko.net & https://blog.senko.net
GitHub: http://github.com/senko
Mail: senko@senko.net
You can already so that with classification models such as ModernBERT, at 0.4B.
Jev's value is its zero shot performance without having to fine-tune.
1. For a techie, you can already de-Google yourself trivially by asking the AI to build you a bespoke web crawler and run it on Lambda or Cloudflare workers to index the entire internet and store it in DuckDB.
2. The answer doesn't actually help the user get back useful results from Google.
3. "Let me know" doesn't scale, and it's not obvious how talking to random internet strangers about their Google problems will increase your startup SEO?
Minecraft clone: https://senko.net/vibecode-bench/2026/voxel-opus-5.5.html (Opus 5.5) vs https://senko.net/vibecode-bench/2026/voxel-fable-5.1.html (Fable 5.1) vs https://senko.net/vibecode-bench/2026/voxel-gpt-6-astra.html (Astra 6)
Warcraft clone: https://senko.net/vibecode-bench/2026/rts-opus-5.5.html (Opus 5.5) vs https://senko.net/vibecode-bench/2026/rts-fable-5.1.html (Fable 5.1) vs https://senko.net/vibecode-bench/2026/rts-gpt-6-astra.html (Astra 6)
The above Opus games took ~45min to generate with the cost between $11 and $14 (per ccusage - I'm on a Max sub). Used from Claude Code with xhigh effort.
Full tests with prompts: https://senko.net/vibecode-bench/
> Now, if you can steer an LLM reasonably well, you can quickly build an MVP that goes well beyond your own understanding of the implementation.
Somewhat agree. Five years ago you could build an MVP without understanding how to open TCP sockets or how to parse HTTP headers. You didn't need to understand relational databases, let alone B-trees or cache locality. You didn't need to know how to install Linux.
Now you don't need to understand the details of connecting to Stripe or Auth0 or setting up a Kubernetes cluster.
> You have to intentionally build your understanding as you build your projects.
Some things you need to understand-others, not so much. Depends on what you're doing, the scale, risks, etc, but that's always been the case.
If BS slightly raise the price and ruthlessly cut costs (think - leave just a skeleton crew to keep it running), the calculation may make sense even if there's slow churn in the users and no special lock in.
(disclaimer: pure outsider speculation on my part)
(this one is older but the other one got to the front page and will accrue more discussion)
Miro acquired my startup back in '21[0], so this is bittersweet as our users will be (eventually) affected by this. Founders and employees at Miro definitely deserve to cash out for their hard work and it's unclear (ie. not publicly known) if they will since this is a 10x down round.
Also weird that they had to sell - they were one of the most established (visual) collaboration platforms out there. I'd be keen on learning about the inside story if it ever gets public.
BS is known for aggressive management of companies they acquire, optimizing for cash flow and revenue, firing most/all of the original team, etc. Their acquisitions before Miro include Airtable, Eventbrite, AOL, Vimeo, Meetup, Evernote ... for at least a few of those, my direct experience with the product afterwards or in chats with their users, the platforms does get enshittified.
Hope that's not a deal here.
> Wasn’t a PM tool.
I know a lot of people who used it for one or both of these. Also for architecture/system design (ie. not UI - although I guess some used it for user journey or wireframing).
As to why the prompts aren't shared, for these more complex things it's most likely a somewhat messy process (ie. not a single-prompt one-shot creation; some back & forth) that would make the whole thing seem less spectacular.
Likewise for the end result - it's probably cherry-picked what works well. For example, in Claude models' resuts I always get stuck in water (something about height/jump calculations is off), where with Astra I didn't have that problem. These sorts of issues you can only spot if you try to playthrough yourself.
This is just my speculation tho. As for me, I do the tests because I'm interested in the results (easy comparison across time & models) - then I started sharing them because people asked.
I have OpenAI and Anthropic subscriptions so testing these is not an extra cost for me (I'm sloppy around recording the tokens & API-equivalent cost tho - have to improve on this). For the other models, it's total a few bucks per month or so.
So if you're careful about the cost, it's not too much, especially for a serious youtuber who's doing it for commercial reasons.
Finally regarding your comment about cloning the popular enshittified games - I don't think people are going to be doing that for testing, but if you want to have a different spin (or do a close-enough clone for yourself) on a game you loved, the modern AI systems can often deliver!
[0] from the link you posted "i am in disbelief, this is insane, bro what is this, you can't believe it's ai" - yeah...umm, it's not that good :)
Compare Astra[0] with GPT 5.4[1] which was OpenAI's state of the art just six months ago.
(all tests on more models with code and prompts available here: https://senko.net/vibecode-bench )
Yes, it's not a scientific benchmark but it's a good heuristic.
For a better eval, create a one-page prompt / mini spec related to whatever you're using the LLMs for, and see how well a particular one works for what's important to you.
0: https://senko.net/vibecode-bench/2026/rts-gpt-6-astra.html
Instead, some telecoms give you a generous number of "credits" for calls, transfer, etc, and sometimes try to limit these to your own country. Then you have to pay overage fees (the same in your home country or elsewhere within EU).
The legality (or ToS language around this) is dubious, but the fact that telecoms try to skirt around the rules doesn't mean they don't work - switch to a better provider.
I just spent a week in another EU country and didn't think for a second about data usage, calls, fees, etc.
Combined with not having to do fx or even carry a passport (national ID is enough), minimal queues at the airport and zero customs annoyance, travelling across continent was as much hassle going to the next town (modulo actual travel time).
(Not to mention the actual benefits of membership, in this case, EU institutions slowly but surely cleaning up house).
We all live in our personal info bubbles and sometimes it's hard to find information if you don't actively seek it out. This doesn't mean it doesn't exist.
I think "being as widely-understandable pun as possible" is definitely one of their naming priorities.
Doable but annoying.
Ignoring the political issues, my problem with Twitter is more prosaic: the character limit is too short for basically anything of substance. I really don't like the thread style it effectively forces.
(Oh, and the new name is unbelievably stupid. I ain't never calling it that.)
On LI the majority of the stuff is so bad, I have zero self-awareness and just write whatever comes to mind (ie. an approach more like what you'd do in a comment like this one, than a "proper standalone post"). Many of my posts are just commentary on what's going on in the niche (eg. comments on a new AI model) which doesn't require a ton of extra research.
I also try to have a weekly/monthly cadence: at least one LI post a week, and one blog post on my own domain per month.
With the Elonification of Twitter there has been a splittering of tech/biz audiences across different platforms, each with their own problems, and the split is different in different parts of the world.
It so happens that LI continues to be popular with the audience segment and geographic location I'm primarily working with.
Agree with all negative comments about the platform tho.
I find that just by not being cringe, obviously AI or obviously over the top sales-y makes the posts stand out and raise my (consulting) profile enough to land a meaningful number of clients, just as a second-order effect.
Lately I've been even more annoyed at the walls of its walled garden being raised higher, eg. no way to export my own posts, or having to log in to open linkedin shortened urls, which breaks my firefox site containers workflow.
I have a half-assed bwrap-based solution of my own, but flar seems much more thorough, and I like that it's a single binary with sensible defaults.
Edit to add: instead of "adapt" (which would imply "embrace AI", which I didn't say), my intent was more along the lines of "be adaptable."
In Django you'd typically use simple classes (models or forms, or even dataclasses nowadays) more than dicts everywhere; n+1 is trivially avoidable (as another sibling comment points out, and you also have multiple packages that autodetect such cases if you've missed them).
Python in general has a more "consenting adults" than "defensive programming" attitude (which doesn't mean exessive coupling or spaghetti, but the approach is different from the Java or C# mindset).
There's no one THE correct style of programming.
I've seen the horrors Java devs start doing on a Python project when trying to "fix" things, where by "fix" they mean use patterns they had to in previous gigs.
It's a different world.