emdash

I would say if we were time travelling, $1 slice pizza

It’s not “brothy”, it’s stew.

I want to know about the lives of people who think this is plausible enough to fall for

So, I’ve been told to use these tools at work, and found a use case that I didn’t hate.

Hooray!

I also fed the content, I wasn’t asking it to come up with anything on its own. Here’s the thing: Gen AI is still making shit up. A lot. And unless you’re an expert on the subject, it all sounds pretty plausible. Which means I don’t TRUST IT to work on anything I don’t already know well.

For my use case, the tone of the writing wasn’t supposed to be personal. So the LLM output worked fine here. Other touted use cases like email writing, etc., I hate because without that touch of the personal in your communication, why even bother at all?

I’m not going to play into the stupid hype by saying how much time this saved me. (This was for internal documentation, which I don’t think anyone enjoys writing.) But I have some theories on why this worked well for me, and asking Gen AI to come up with content from just a prompt generally doesn’t.

In the prompt I told it to make the text sound professional, and to not use em dashes (lol). The result was still quite true to my input and intent, but it was polished enough to mostly use as is. My dictated input was very rough - no punctuation, run on sentences, like I said, a brain dump.

God help me, I actually found a case where an LLM has made my job more efficient. I still don’t like the tech or the hype for many reasons.

I used dictation (built in accessibility feature) to brain dump what I wanted to write. Then I ran it through the LLM (MS Copilot) to edit it.

My dream last night mentioned Dole Whip and now I really, really want one.

Last time I travelled through SFO (a week ago) all the billboards were for AI. A few years ago, they were all for blockchain. In a few years they’ll be for something else.

If SV wants something to stick they should try solving real problems, not trying to shove unasked for “solutions” at us.

I’m not convinced it’s foil, it looks like it might be a gradient applied to the lettering. Especially if you consider it might’ve originally been “designed” on a single line.

Seeing as he’s “decorated” the office with plastic bits from temu spray painted with gold coloured paint…

Between the gloomy weather and the nonsense clock change, it is dark outside already at 3:30 PM and I hate it to an unbelievable degree.

Update now that I am home: the tiki bar had room at the bar and was excellent, I didn’t have to wait in line for the donuts and they gave me one free. 10/10 would visit the terrifying crime ridden hellhole again.

I really loved the shitty little pyro I created for Treasures… turns out gnomes are fun!

Sports gambling and AI are two of the modern 4 Horsemen of the Apocalypse.

I currently have a lot to drink about.

The video continues with an example that the agent could autonomously call police when the camera sees a crime. Seeing as its alerts are triggered by the dog next door but didn’t go off when people stole an Amazon package right in front of the door, yeah, no.

I’m using online learning resources for agentic AI at work. Some of its examples are less convincing than others. Why yes, I do own a smart camera doorbell. It is accurate less than HALF the time and misses my damn package deliveries. Not off to a great start here.

In a couple of weeks I’m headed to the terrifying city of Portland, which is full of riots and fire and looting and so on. Gonna try to get into the good tiki bar and wait in line for too long for donuts. Maybe I’ll eat from a food truck. Gonna be dangerous but I think I can do it.

You have a stuffy that looks like your cat!!!!!!!!!1

It's your world.

Take it back.

A D&D campaign but your characters are based on RV brand names.

About 8-9K employees. We are so slow to ship anything it’s painful.

Likely no one will ever see this issue. Common sense would say, ignore this one, it’s not major. But common sense is not common in large orgs, this little scrollbar bug has already wasted 5 people in a half hour meeting to discuss it, and likely a bunch of paperwork for an exception to ship with it.

But! The end users are not using this resolution. Most are on huge desktop monitors. I have no real telemetry, just observations from site visits, but I’ve actually NEVER seen a monitor this size in use. For laptops and desktops, there is no scroll bar.

The process to change this target resolution also means changing all the software if it doesn’t fit the new size, which no one will approve, so we’re stuck with it. At 1280 pixels wide, one part of the display has an unwanted scroll bar. So it fails. And we can’t fix it easily.

Large organizations get so bogged down in process and checklists that they lose sight of reality. An example from product design at my current job: a long long time ago someone decided the target resolution for all our web software was 1280 by 1024. QA tests and files bugs based on this.

Now that I own a home I must maintain, I have come to the belief that both a) indoor plumbing and b) climate control were HUGE mistakes. SURE let’s put impossible to access pipes of toxic crud all through your house what could go wrong

Best flavour