In a normal world there'd be a market research phase and the tech CEO would be looking at topline stats on where they think users are looking. And then they'll develop products to meet their requirements before competitors do.

Instead we have a made up product (that other AI platforms may offer for free) and a market research report that is telling the CEO that consumers consider the product 'slop'. And the result is a brand deal, money put down, and the CEO having to try to convince consumers it's not slop before it even launches.

This is what we call 'dead on arrival' right?

Check out https://www.youtube.com/@noisygroup JOEY on YouTube to see how it should be done. Their work still fails a lot of sense checks but it's orders of magnitude better than the slop your $60 got you.

Your attempt is full of visual inconsistencies like the architect's drawn line behaving like taut string as it's drawn (and being a pointless shape over the blueprint). The safe handle wheel turning by itself whilst the guy smiles idiotically at the camera like a 70s comedy show. The reviewer reviewing absolutely nothing on multiple screens at the same time. The orchestrator doing pretty much the same thing but with DNA helixes?!?

The overall message and feeling derived from the video is - sloppy, cheap, shallow, as if produced by a company promoting a product that will be all those things.

People are going to generate some pretty effective media from this tech eventually but there's a sliding scale from free slop to million dollar ad equivalents. And yours is closer to the free slop end of the scale, closer than your $60 budget would even suggest.

538, named for the combined number of senators, representatives, and electors in the US, was a political journalism site. It used polls and data to ground its articles and opinions. Largely helmed by Nate Silver for a large portion of its run, until he left.

It got 49/50 states correct in the 2008 presidential election which catapulted it to fame. It never scored that accurately again but it's data driven charts and tables were a great pulse point for US politics. You could see recent aggregated polling data day by day in the run up to an election. And you could read analyses of polling data and trends.

There's nothing like it anymore, it didn't make money and didn't make spot on predictions reliably. It didn't need to do either, it was a good pool of data at the end of the day.

Everyone seems to rely on Polymarket now, which is a much more cynical way to aggregate polling data - by betting on predictions.

Within technical work most of the interfaces I deal with are high contrast, and do not rely on colour in a way that would lead me to get things wrong. There are some issues with documents containing black and dark red text that I can miss. That's solved by telling the author not to mark important changes/highlights in dark red. There's a tracked changes function, there's a highlight function. I am covered.

The more technical my work gets the more I tend to be covered by WCAG 2.2 guidelines. I don't have to do anything based on colour, or report colours during a task often if ever. So as long as I can see the contrast between things I am fine, they can be whatever colour they want.

One thing on colour correction though, you don't have a colourblind person's 'profile'. In fact really, any given healthy person's profile of colour vision is not going to be the same as anothers. A lot of colourblind 'filters' in things like games and apps are barely serviceable. When I use them I find the colour profile I am used to in the wider world flipped, and the semantic meanings given to colours, or their hierarchies, completely changed. Within a UI that often isn't helpful.

My colour blindness causes some frequencies to bleed into each other in my perception of them. That means I struggle in some degree of overlap of red and green to pick out or identify that colour. So make the colour purer for me. Do that with contrast, and saturation. Don't flip it to a set profile or palette. Don't make STOP signs Cyan instead of Red. I CAN see Red, sometimes I struggle when a Green comes along on the edge of its spectral profile and triggers my red cone the same way a similarly fringe red frequency might. But pure Red? Easy, I know that one. Don't do anything to pure (primary) RGB values.

And note that computer monitors are doing all this with RGB, in their own available gamut and brightness/contrast settings, possibly with OS-based HDR setting son or off.

You can try to make a perfect system for every variation but the end user won't see it as precisely as you intend.

I'm married to someone I went to board game meetups with before we dated. I'm very good friends with several people from that group now. We play board games periodically still despite the group no longer being active, and our reasons to meet up are now more about socialising than board games. Board games was a brilliant icebreaker, but it took effort from everyone involved.

Perhaps board games are not the best socialisation method, but I'd argue against that too. It's a very good way to get like-minded people in a room, sitting down, and collaborating.

However, "I know many people who say they’ve tried a couple of board games like Catan and really enjoyed it, and are surprised to hear that I think it is a bad game." is spot on and the author cannot be faulted for that opinion. I'd rather play Russian Roulette than Catan.

I don't see any threats, I likely missed them in my scan down the sections, but it does just seem like it tried to meet you in the middle and its output flew over your head as it miscalculated the bro-speak conspiracy theory level you were at.

"honestly i know where the next "revolution" should be happening. those who get it, get it.

let those who dont waste their money

bye"

Hilarious way to speak to Gemini. It's going to spew that right back at you.

This does nothing to stop the rise of dairy alternatives, nor should it. Nobody is going to stop asking for oat/soy/nut milk when putting in their drink order, just because corporations can't use the term. It's common parlance now, and using 'milk' to refer to a non-dairy liquid has been done for hundreds of years at this point.

This is such a strange decision.

They're no longer a best of both worlds, good overall range, not as bad for the environment.

They're now worst of both worlds, bad electric range, bad for the enviroment.

Anyone considering an EV isn't going to want to compromise on those two things as much as they used to.

I wouldn't trust any tool that can't spellcheck their own website.

However, it's not GTM's fault, at all. GTM gives marketers a very powerful suite of tools that they normally have to rely on devs for. In fact it gives them site-breaking power.

But it's not GTM's fault a marketer can't track a button click in the scenarios. The marketer just needs to push the responsibility back to the dev who made the crummy site. GTM can't help if a site is built in an opaque manner. GTM can't help if a marketer can't learn to use it competently either.

Depending on the volume (100k is a lot) a human check may be made and result in a manual action against the site which would lower discoverability.

Discoverability correlates with link velocity, if velocity is uneven it is suspicious and punished.

It would help crawling and indexing technically but that's something you can do for free via Google Search Console with sitemap submission and manual index requests.

Automated backlink creation is and always will be a 100% bad idea in the short and long term. It is a waste of money and effort and only every something scammers and bad product managers endorse for their own benefit. Any positive effect always turns out to be misattributed.

I'm referring to the Act's powers to compel companies to actually do things in my original comment. I don't know exactly when various parts came into effect that would constitute that, but for the point of my post I'm going on Peter Kyle's own website's dated reference to holding companies accountable.

"As of the 24th July 2025, platforms now have a legal duty to protect children"

https://www.peterkyle.co.uk/blog/2025/07/25/online-safety-ac...

I don't understand why people are taking issue with that. Peter Kyle is the minister who delivered the measures from the bill that a lot of people are angry about and this latest issue on X is just another red flag that the bill is poorly worded and thought out putting too much emphasis on ID age checks for citizens than actually stopping any abuse. Peter Kyle is the one who called out objections to the bill as being on the "side of predators". Peter Kyle is now the one, despite having moved department, who is somehow commenting on this issue.

Totally happy to call out the Tories, and prior ministers who worked on the Bill/Act but Kyle implemented it, made reckless comments about, and now is trying to look proactive about an issue it covers that it's so ineffectively applying to.

Ironic that the minister who spearheaded that awful bill (Peter Kyle) as Tech minister is now being the government spokesperson for this debacle as Business Minister. The UK needs someone who knows how tech and business works to tackle this, and that's not Peter Kyle.

A platform suspension in the UK should have been swift, with clear terms for how X can be reinstated. As much as it appears Musk is doubling down on letting Grok produce CSAM as some form of free speech, the UK government should treat it as a limited breach or bug that the vendor needs to resolve, whilst taking action to block the site causing harm until they've fixed it.

Letting X and Grok continue to do harm, and get free PR, is just the worst case scenario for all involved.

TITLE: More Than 60% of Consumers Now Start Daily Tasks With AI

2 statements from the article:

1) "More than 60% of consumers used a dedicated AI platform in the past year, signaling that AI has crossed into mainstream behavior rather than niche experimentation."

2) "Over one-third of Gen Z consumers and Power Users now turn to AI first when starting personal tasks, marking a clear move away from traditional search and browsing."

The 60% stat refers to having used AI at least once over a 12 month period. The personal task stat is for gen Z and power Users and may not even be 'daily' tasks.

Wrong. As the article even admits, multimodal AI turns imagery into text values for analysis (or even tokens/vectors). Sure, current AI bot crawlers are not multimodal for speed and cost reasons. But stating the future will be text based when we can already do visual processing with AI is just shortsighted.

"design should no longer be reserved only for humans, but also accommodate AI agents"

The design of the web has never been reserved for only humans. It's always been machine readable. In fact design principles for SEO have specifically integrated machine-readable features like alt tagging, schema, screenreader tech etc. What has been true is that various organisations (google included) have spearheaded a user-first approach. We can instead predict that this will follow with AI, lead with human first design and have the AI mimic to understand content. Since AI is far more human-like than your standard SEO crawler this principle follows even more strongly. Why prioritise AI design over humans when AI is trying to behave like we do?

"shifting from a “user-first” to an “agent-first” mindset"

Not sure I understand this section, it suggests making 2 distinct interfaces, one for humans, one for AI. As if APIs don't already exist? Plus, for when an AI is being agentic, they absolutely should walk through the same steps as a human. There is zero reason to duplicate an interface for a tool that impersonates a human, that just multplies all test cases by 2. Pointless waste.

"In the future, the user experience will focus primarily on supervising and verifying the actions of the AI agent."

What? UX does not supervise or verify, UX itself is not an agent like a human is or AI can be.

And then the article actually gets to sensible technical UX advice for SEO and accessibility that will also benefit AI. But it was mostly slop to get there. Was this written by AI?