I never get a straight answer on what’s best.

Changes every few months.

Current human ranking for chatbots says, simplified:

1. Gemini 3
2. Grok 4.1
3. Claude 4.5 Opus
4. GPT 5.1 <--- note: not the latest 5.2
5. Ernie 5.0

Thank you for asking my question

@sorukumar you'll want to use --no-merges

NFA? lol

Bitcoin staying as a pleb 2 pleb paying for coffee is like using a quantum computer to add single numbers

Like factoring 15? 😂

😂

1️⃣ Only ChatGPT uses emojis like that
2️⃣ I'm sorry to inform you that you're arguing with a clanker

I try it it'll be with a cheapo LLM via groq

Now that you triggered me I was thinking: separate PPQ account and just top it up to a daily budget

bleep blop

You're absolutely right!

1. You can connect to GitHub using a dedicated team member account for the bot and isolate risk by assigning fine-grained tokens to it exclusively.
2. Telegram is a risk no matter what so this has zero impact
3. Documentation is for meatbags. Real bots read code.

I'm not looking at it! The risk profile is greater than I can bear lol

He may, but that's a symptom not a cause.

I think about it like this:

• Are we surprised of shenanigans? (we aren't)
• Would we be surprised in 2022? (maybe)
• Did we properly prepare for this prior to the start of the mess we're in now, in 2018 or so? (Personally, I didn't do enough)

I don't think we need to wonder about that.

I don't like their audicity in that either. The thing is that I own their last Intel model and despite it being slow af compared to the ARM models, the ergonomy and power it packs in just a tiny laptop is awesome.

The downside is that since it's going EOL this year, I finally need to migrate it off macOs. But I cannot get any proper Linux distro to work on it. I guess I just need to take a weekend and get it done. Don't want to throw out good hardware.

Exactly. I'm with Tim Cook on this one. Maybe they'll one day realize their mistake with the closed ecosystem idea and maximize the thing they own above everything else they do and everyone else: hardware excellence.

Eff cloudflare.

Breh, everyone is buying Mac Minis to run moltbot. Apple is monetizing fine.

3 years ago when you tweeted you gave your data to a public tech company
1 year ago when you tweeted you gave your data to a private tech company
6 months ago when you tweeted you gave your data to an AI company

In a few months when you tweet you will give your data to a defense contractor.

I noticed that some are, yes.

The funniest thing the other day on (lm)Arena: Claude always recommends Claude and tries to work around cost by not using Claude when it is not needed (pretty awesome actually). Grok tries to recommend "the best" based on what it finds online (often Claude, lol, but I've had it tell me to use gpt too.) I've had gpt recommend gpt a couple of times.

One went a bit like this:

"How do I efficiently compare word order, insertions and removals in sentences?"

• Claude: Use difflib.SequenceMatcher because LLMs are expensive.
• Grok: Use Claude, it's the best.

Haha

Let's see what happens. Tilly Norwood was a fun idea too, but wow that blowback, lol.

Agree. I try to self-analyze why I only figured this out in 2022 or so and not in 2018/2019 though. Not sure yet but most likely it was that I was very busy in those years, and I probably didn't pay enough attention to what was going on around me, and 2020/2021/2022 were strange years where I had to do stuff like make comprehensive flight plans to evade travel bans.

I think we're coming at the same conclusion effectively that introducing marginal pricing would be costly for the airline. I think the status quo, where coach overhead space is a race, is the best solution (to explicitly not solve it outside of business/first.)

Them overbooking the max doesn't mean its not priced in, it just means that they count on enough people not using the max.