belval

Interesting. I did try Opus 5.5 but it didn't generate models that were as good. A lot of clipping/empty paths. I had it create a drone, the drone frame had very obvious defects where some pieces weren't attached to the frame.

Perhaps my approach was too one-shot, do you have a harness that creates model reviewer/refiner?

People have been talking about Sol 6.1 and maybe for coding they are equivalent, but Astra 6 is still miles ahead on just about everything else.

I had Astra agents build 3D models and gave them direct access to my 3D printer. The quality of their designs is much better than what Sol has been able to make for me. Enough so that I think Astra is a slightly different architecture/encoder, not just a bigger model.

Picture, to get a sense of it: https://github.com/Belval/crossflow-fan/blob/main/renders/ca.... This printed without any obvious defects (yes the fan blows air)

I'll go against the grain and say that I like this, I genuinely don't think this is anything but some internal team being excited about their product and giving you some firmware to connect stuff to their agents. Meta the trillion dollar company will not move at all from this.

I played it in VR and it still does.

No Man's Sky suffers from planets being "deserts", they have different attributes but it's still a fairly lifeless empty world.

Ironically this is realistic but also not terribly fun as a player.

I think it's the:

> a first for the open-ear form factor

That is being called out as comical.

Gastown lingo: https://github.com/gastownhall/gastown

As a fervent Claude Code user who made the switch to GPT 5.6 Sol over Opus 5 over hard-to-read prose this makes me happy. I love your product but the current models are very hard to work with if you need to do a lot of context switching. Brevity is key.

I really don't like Wayland, I recently (last Friday) had to revert back to X because somehow they messed up enough that in Zoom screensharing during a meeting with XWayland can crash my audio drivers.

At this point I don't know who it is for, but it's not for users.

Working on a tool that uses their model and the Opus models are all guilty of this.

"Please run N tasks doing X for Y duration, I want to test something"

"Hmm but that would waste tokens, we shouldn't do this test"

What the hell type of product is that? I wasn't asking, I was ordering it to do something, if Anthropic can't get their model to do as I say then I'll just go use someone else's.

Sure but the GPUs in the article are from 2021 and cost ~500-600$ a pop. The OP bought 4 so that's 2k, let's say another 500$ for the rest of the chassis and you stand at $2.5k for a monster server that will burn electricity and warm your house.

Compare that to a $5k DGX than consumes much less and has the same amount of VRAM and there is a real question as whether this is worth doing at all (well aside from the cool factor).

More than half imo, my Forerunner 265S is pretty much a full-blown coach that fits on my wrist. It adapts training to my sleep (such a great feature when you've just had a baby) keeps track of load, low aerobic, high aerobic, etc...

It's been a fragile piece of hardware, I am on my 3rd one in as many years (always replaced under warranty though), but I still probably wouldn't switch to anything else. When it breaks out of warranty I'll just buy another one.

Someone instantly started spamming the bee movie's introduction. Solid pun.

> Would they not be in favor of _less_ regulation so they can provide services to their users with fewer legal considerations?

Regulatory capture. If the handling of user messages requires constant scanning and there are enough rules that you need a team of lawyers, then only Google, Meta and Apple will be able to afford it.

The 42.7M tourists that went there in 2025 just enjoyed seeing the incredible decay of a dying country, a morbid fascination or sorts. /s (obviously)

It's so sad that Disney burned some Star Wars fans so much that they will never give Andor a chance. It is such a work of art.

You still have near-infinite tokens, you just don't get awards for using a lot.

> Do you think ai usage will decrease due to this change ?

Frivolous usage sure, internally at Amazon there is a subculture (if you can call it that) of award chasers. Using Kiro for mundane task to burn tokens does not sound that far-fetched.

Overall usage though no I don't think so, these tools have some pretty wide adoption at this point and not by people chasing awards.

This is how I learn that they shutdown the awards for Kiro usage.

But more seriously, while this is worded as a result of abuse, it had been up for more than a year at this point as a way to push people to use Kiro (you get badges for each "level" basically). Once you reach a point where everyone is using those tools, it makes no sense to keep it around.

Also it was not related to any performance metric, it was a pure vanity thing of getting virtual awards to display.

Perhaps not a monster, but in "It takes two", there is a particular scene where you have to murder a friendly stuffed elephant to get your in-game daughter to cry.

I have been playing video games for decades at this point but that one really shook me up. You pretty much execute a toy begging for its life. As soon as that scene was over it genuinely took me a few days to come back to it.

That was a hard rewatch: https://youtu.be/12FNU8bNEbE?si=BKZCynsHhoz5GN2m&t=65

> Smoke-free

> Less than 5%

Incredible milestone but when 1 in 20 is still smoking I feel like it's a bit early to call yourself smoke free.

Maybe, but I have trouble with the framing. Referendum votes are >50%. If a foreign nation can get >50% of the Albertans to agree to something, that's still democracy.

Yes it feels wrong for the US to be giving money to influencers to influence the vote, but it's not like those voters are being coerced. In their opinion, Alberta would be better as a separate country.

Whether that opinion is enlightened or not has no bearing on it being democratic or not.

Free mulligan means on your third hand you would only bottom 1, not 2.

Besides, with a 37 lands commander deck, the chances that in 4 hands (if your limit is bottoming 2, 5 cards in hand) starts getting low. Around 1 in 10 and that's ignoring mana rocks (2 lands + 1 mana rock) which probably brings it much closer to 1 in 14+.

If you add the chances of more than 4 lands (also fairly undesirable) your odds of a bad forced hand climb to 1 in 7, which is still pretty far from the 50% chance the original post is talking about.

Without the free mulligan you would be at 1 in 4, which is why I said free mulligan is the way to go.

Alternatively you give everyone a free mulligan and if they decide to start with <3 lands in their hand or no mana ramp that's on them.

My issue with guaranteed lands is that they remove the randomness of some lands. I am not guaranteed to get my "no maximum hand size" land and my rogue passage to make my creature unblockable in a typical game of commander. I have to plan for it by using stuff like expedition map.

> No Land destruction. It sucks to be unable to cast spells.

I have yet to find someone actually running land destruction in their deck, it's such a hated mechanic.

> selling this at $15/hour/user??? That math does not math. A quick google says there are between 1.5 and 4.4 million developers in the US alone, let's say it's 5 million, to be generous, and each of them is subbed to this for 8 hours per day, continuously. That's 600 million per year in revenue

That math is not mathing. $15/hour/user, with 5M devs, 8hrs and 240 working days per year that is 144B in revenue.

What exactly should the author have done differently? It's part of the leadership roles to understand the power structures within your organization. Reading between the lines, a new team was thrusted onto an arguably functioning sub-org to address concerns that they had not themselves raised. Then the expectation was for that sub-org to take a hit on their KPIs to onboard to the new teams platform.

It's not "tribal" to refuse to do something that is misaligned with all your explicit incentives. Otherwise we'd have to pay lip service to every internal tooling team just because they exist. It's the leadership team's job to keep pushing if they strongly believe the sub-org leader is acting in bad faith.

For me the next step should explore how to cut out the firing part of the process altogether, pottery looks cool but the process requires a lot of energy. Perhaps it could be done on a piece of wood planed by hand? You can get those fairly flat. Then use copper tape (or laminate your own copper really) with some homemade adhesive?

Actually now that I think about it you could just make pine rosin (pine resin + alcohol) as your adhesive. For the copper laminate this might be harder without steel rollers or a way to cut.

Bah the other wheels usually die of natural causes if you wait a bit.

I am a big believer in Amazon's "1 > 2 > 0" for "one perfect solution is better than two, but both are much better than no solutions". It's also misunderstanding how Amazon works. If you want to move fast, you can't wait for an SVP 6 levels above you to approve every effort. Instead you build something and then you run it up the chain to have it be adopted at team/org-level.

I didn't say it was devoid of substance, the poll part is actually interesting (and worth discussing!) it's just that it actually appears *after* the sloppy tweets and "someone pretended to shoot at Sam Altman's house" screenshot as if that was somehow relevant.

This is poor reporting, almost needs a checklist:

[X] Tweets and instagram comments presented as "what society is thinking"

[X] Ties Luigi Mangione and the California warehouse fire to Gen Z discontent (about AI?).

[X] Statistics being used to support the title with little to no regards to continuity: "those respondents who said that AI makes them “nervous” grew from 50% to 52% during the same period" => percentage was 52% in 2023, 50% in 2024 and 52% in 2025, seems mostly flat to me, with the real jump being in 2022-2023 with 39%.