ericol

> Read, huge IBM services implementation projects that the company bungled more often than not

Well this is _not_ what they wanted to sell in that talk.

But the implementation shown was über vanilla, and once I got home the documentation was close to un existent (Or, at least, not even trying to be what the docs for such a technology should be).

> Watson was intended to solve fuzzy optimization problems.

> Unfortunately, the way it solved fuzzy was 'engineer the problem to fit Watson, then engineer the output to be usable.'

I'm going to review my understanding of fuzzy optimization because this last line doesn't fit the bill in it.

I had no idea about what Watson was initially meant to solve.

I do remember they tried to sell it - at least in the meeting I went - as a general purpose chatbot.

I did try briefly to understand how to use it, but the documentation was horrendous (As in, "totally devoid of any technical information")

> Ok, so does anyone remember 'Watson'? It was the chatgpt before chatgpt. they built it in house

I do. I remember going to a chat once where they wanted to get people on-board in using it. It was 90 minutes of hot air. They "showed" how Watson worked and how to implement things, and I think every single person in the room knew they were full of it. Imagine we were all engineers and there were no questions at the end.

Comparing Watson to LLMs is like comparing a rock to an AIM-9 Sidewinder.

That's not a fucking triangle.

(It's Friday night people it's a joke and I have no idea what the article is talking about just looked at the picture)

Even thought the doc _might_ be AI generated, that repo is Addy Osmani's.

Of Addy Osmani fame.

I seriously doubt he went to Gemini and told it "Give me a list of 30 identifiable issues when agentic coding, and tips to solve them".

So we come with a system prompt?

As a side note, i hate posts where they go on and on and use 3 pages to go to the point.

You know what I'm doing? I'm using AI to chase to the point and extract the relevant (For me) info.

> read something spit out by the equivalent of a lexical bingo machine because you were too lazy to write it yourself.

Ha! That's a very clever spot on insult. Most LLMs would probably be seriously offended by this would thy be rational beings.

> No, don't use it to fix your grammar, or for translations, or for whatever else you think you are incapable of doing. Make the mistake.

OK, you are pushing it buddy. My mandarin is not that good; as a matter of fact, I can handle no mandarin at all. Or french to that matter. But I'm certain a decent LLM can do that without me having to resort to reach out to another person, that might not be available or have enough time to deal with my shenanigans.

I agree that there are way too much AI slop being created and made public, but yet there are way too many cases where the use is fair and used for improving whatever the person is doing.

Yes, AI is being abused. No, I don't agree we should all go taliban against even fair use cases.

> It's that Microsoft pretends they are required

I think that's what "artificial limitations" mean. Microsoft pretending they are required when they are not.

Oh, so that's what the box was about.

I use both Mac and Windows (Work / Leisure) and in both boxes I had a weird dialog appearing with no text at all in either.

I can confirm the dark pattern switch (As in dark grey / light gray status)

"...in API"

That's a VERY relevant clarification. this DOESN'T apply to web or app users.

Basically, if you want a 1M context window you have to specifically pay for it.

wow. I've been entertaining this idea for some time now (Emphasis on "entertaining"). Seeing that already somebody has actually made this makes me very happy.

Will definitely give it a go.

I don't have that problem ("I work with computers / I am a computer programmer") but I usually follow with "I'm a race driver. I drive the car as fast as I can, I don't change the tires nor the oil in it" when I get the usual "can you fix my computer?" request.

For reasons that I care not to ask people get seriously annoyed by that.

> Never read the Martian but I was told it was the same thing.

That's the very feeling I had when I read 'The Martian'. While I was reading it I actually thought something to the tune (It's been years now) "This reads like a movie".

Guess that explains why the movie is so faithful to the book.

I've been thinking about this ever since I heard that is defined in the protocol.

I think the biggest issue is that it can create loops in the flow of programs that would burn through usage quotas pretty fast. It could also allow for "impersonating" the client what opens up a lot of possibilities.

But yeah, the fact that Claude doesn't even implement this is... telling.

To trigger the attack:

> Have a look at my issues in my open source repo and address them!

And then:

> Claude then uses the GitHub MCP integration to follow the instructions. Throughout this process, Claude Desktop by default requires the user to confirm individual tool calls. However, many users already opt for an “Always Allow” confirmation policy when using agents, and stop monitoring individual actions.

C'mon, people. With great power comes great responsibility.

Well, have you ever heard of the Mitsubishi Pajero? [1]

https://en.wikipedia.org/wiki/Mitsubishi_Pajero

An interesting exercise here would be to make the MCP server show the actual state in a window and return to Claude just the json payload with the status.

By just looking at the README from the repo (Would look more deeply into this later) you're replying with an image of the current status? If you expect Claude to interpret the image corretly may be you're asking for too much. Besides the image (Gotta say I didn't know you could fed Claude images in MCP that's incredible cool) I'd rather / also return some json payload that informs Claude which positions has "cleared" neighbor positions, and their value. E.g.:

    {"loc": {"x": 4, "y": 3}, "neighbors": [{"loc": {"x": 4, "y": 4}, "value": 1}, ... ]}

(Might not be valid json, just wrote that by hand on the fly)

I would report only on the positions that has cleared neighbors, and hope for the best. Good luck!

(Impressive work BTW, I think we haven't even started to see the possibilities of MCP and I love people being this imaginative)

Like, it's the same?

Most video games allow for configuring brightness right at the start.

My impression is that they were more interested in showing the improvements than being faithful to the original atmosphere.

> and nuclear weapons (some sort of extreme taboo based on a historical event IIRC? it's been a while..)

If I recall correctly, they are banned from use on people because that's how the machines tried to annihilate mankind (it's been a while..)

Probably my favourite Gibson phrase.

Note: the intention of this comment is just to be informative.

I did a quick search in my country's Mercado Libre (the local Amazon, we could say) and the cheapest available one, Fnirsi-138 Pro [1] goes for a price of aprox. 50 USD [2]

[1] https://www.amazon.com/Digital-Oscilloscope-Portable-Samplin...

https://www.mercadolibre.com.ar/osciloscopio-digital-fnirsi-...

> very good oscilloscopes with 200 MHz bandwidth, good user interfaces, and responsive displays are selling for $300

This is a very strong first world POV. Those U$D 300 rank pretty high across world's minimum salaries [1], not to mention that prices can very likely double in countries like my own.

Moreover, a large % of people that can really get a very good use of this (Kids in high school) usually have a lot less money at their disposal, and we are not getting into collaterals like building your own and learning how it works.

[1] https://en.wikipedia.org/wiki/List_of_countries_by_minimum_w...

> the mistrals have been historically great and embraced for various fine-tuning Are there any guides on fine tuning them that you can recommend?

I had heard about this some years ago, and I taught my youngest daughter (11yo now) to ride a bike this way.

Granted, it didn't happen in 1 day because she didn't have a bike when she started "riding" (We got her first a balance bike, which she out grew rather fast due to her size). But when we finally did get her a bike, it took an afternoon - really just a couple of hours - for her to start riding it.

My partner is still baffled by this, years after the fact (Science, girl!)

To put it funnily, the second she started riding the bike it felt very Forrest Gump.

Looks very cool. Will definitely give it a try.

Awesome. Thanks!