jmfldn

I'm somewhere in between. I'm excited about building more things faster and extending my capabilities. But I also love thinking about the underlying language, runtime, algorithms, the wider system. I want LLMs to enhance this for me, I want my understanding to go up as I write less code. It's also key to my job as a lead that I maintain understanding of the system for debugging, security etc.

So if I can do both with these tools, then great. I want to cognitively offload in a way that allows me to focus on the important bits. And I'm writing instructions to the LLM to help me do that eg 'help teach me this bit'. A builder and tutor at once.

I understand he’s making a technical point about efficiency, but language isn't neutral and I think it betrays something deeper. It's such a glib and shallow point too that I think it should be called out since he has a track record of saying some incredibly shallow things about AI, people, politics, and everything really.

This is a profound category error. What Altman reduces to a 20-year 'training' cycle fueled by 'energy' is what we, in the actual world, call life. It is a stunningly hollow perspective that uses the language of industrial output to describe the human experience. While he is likely being provocative to keep his product at the center of the cultural conversation, it probably exposes something about him.

Well exactly ;)

"In 2025 finally almost everybody stopped saying so."

I haven't.

If it moves and connects with you then it's real music.

It's fine to have a preference for live musicianship, but the 'real music' argument has been leveled against every new musical technology (remember the furore around Dylan going electric?). It dismisses contemporary creativity based on a traditionalist bias that elevates one form of execution above all others. There's also a huge amount of skill in producing good electronic music. It's always hard to make good music no matter the means.

Seeing all the experienced engineers on this thread who feel discouraged is really depressing.

Not because I disagree with you, I don't! It's because I fear a gradual brain drain of people who actually love their craft and know how to build things properly. I fear we'll end up with worse software thats simply 'good enough', built a atop a pile of AI slop.

If it's cheaper but with acceptably worse results, I fear this is good enough for a lot of companies.

Sure maybe you're right. I'm just so underwhelmed by what I see in my day job that it's hard to map error prone and limited deep learning tools to what is being described here.

I don't see a strong argument here, more just a hope that something will spark this sci fi trajectory you describe. I'm sure big enough changes are afoot, but I think that the AI we have now will turn out to be much more of a 'normal' technology than most people expect.

A fun read, but wildly implausible. Perhaps there are other frontier technologies out there that get us even a fraction of this. But if we're talking this time horizon, I assume we mean LLMs or some other related thing? Are you joking?

I love his attitude, an example to us all. Woz knows what's important.

Best compromise is a markdown file. You can read with it with Obsidian if you want a better gui, but you can also just treat it like a simple text file if you prefer. No lock-in to an app.

I agree that complex todo apps are a bit of a waste of time.

Less training data for Scala is the issue probably. Imagine how much Python is out there.

Depends what we mean by better. If you prefer rock music to Bach then great. Enjoy! I love popular music and classical for different reasons

But if we're talking skill, intellectual depth, craft, then there are objective criteria. Take Bach, his music is like a masterpiece of engineering with its unparalleled compositional complexity and craftsmanship. His mastery of counterpoint being but one example. His work represents a pinnacle of musical architecture, establishing foundational principles that profoundly influenced centuries of Western music.

That just doesn't compare to most pop music does it?

I think this is a bad take by the way. Sure, coding is changing, but people still need to understand AI code. And it's also getting boring pointing out that coding is only part of the job. Often the easiest part.

Impossible without either harming your own wellbeing, the child's or both. And I suspect the quality of your work, even if you somehow manage to somehow pull it off, would be very badly affected. Caring for young children, let alone babies is an intense, full time job.

Technically breathtaking, but why do these examples of AI-generated content always have a cheap clipart vibe about them? So naff and uninspired given the, no doubt, endless potential this technology has.

I also feel a sense of dread too. Imagine the tidal wave of rubbish coming our way. First text, then images and now video can be spewed out in industrial quantities. Will it lead to a better culture? In theory it could, in practice I just feel like we'll be deluged with exponentially more mediocre "content" .

Agreed, it is a bit complex.

I've looked into KaiOS, but I read that its future looks murky at best. In principle though, I like this idea.

https://www.reddit.com/r/KaiOS/comments/15wn0co/is_kaios_dea...

Does anyone here use both a dumbphone AND a smartphone?

I was thinking I could call forward from my smartphone to my dumbphone when I'm leaving the former at home. Seems like a good compromise?

Worse than useless for non-trivial tasks. Great for simple code / scripting questions where I would have used Google in the past. It's basically by Bash script assistant now.

They might be wrong. And since they might be, and we have a chance to mitigate things somewhat, doom is not a useful response. It is paralysing.

Imagine fighting a war. If you and your fellow troops believe you will definitely lose, what happens to morale and the outcome? Morale collapses and you create a self-fulfilling prophecy.

The situation is serious but extreme doom predictions are the outlier in all credible models.

I'm really glad to hear that you found a way to deal with this. As someone who has dealt with generalised anxiety / panic disorder, it always deeply heartens me to read stories of people who have found some relief.

Similarly, I have made a lot of progress through lifestyle changes, but also through reframing. In my case, I try to not worry about the worry, and instead experience it as transient physical and mental sensations (which it is of course). I also find that an attitude of compassion and kindness helps, to direct that feeling to myself. It's something I picked up for Buddhist practice, but compassion practices can be a very very powerful antidote to fear and anxiety for anyone. Generally, the approach is not just passive mindfulness, but active in the sense of actively trying to cultivate a counteracting emotion in the mind / body.

So I guess this is somewhat related to what you're saying. I sort of reprogrammed myself.

I use a Thinkpad T460s (Linux), 4 year old Android phone and an old Garmin running watch I bought off ebay.

The simple fact for me is that all of these types of devices are thoroughly comoditised now, and almost perfected, so even old tech runs well, for my needs at least. I don't experience any issues not being "up to date" and it feels satisfying to not have blown thousands trying to keep up.

I love this take on it, I've watched it a few times.

It goes to show that when you have writing and acting of this quality you don't need any other embellishments.

Ha yes! We're not there yet with this tech, that's for sure.

Yep for sure. Loads to consider here! My initial comment was just a brainstorm /bluesky thinking kind of idea. Clearly the implications are incredibly deep for security, data retention and a host of other concerns.

Perhaps? I'm an LLM / NN layperson so I know basically zero. In my mind, the data is all smashed into a trillion numbers so there's no notion way of removing it as such.

Your point about fine-tuning it out is interesting though. As you say though, you'd have to be really sure you'd really erased it or made it "unreachable" .

An excellent point around data retention and training data. It's not like you can scrub out bits of the neural network weights. I guess such a product would have to factor this in.

Indeed it would. But then the source training data would be bad to lose too. That already exists. I understand it would be a higher level of damage to lose the model of course, but the benefits outweigh the risks, and these could be managed via a very security-hardened product.

Perhaps such products are in the works, but I would love a private company LLM that slurps all messages, docs, meetings, and becomes an oracle that can be relied upon. Perhaps it could even infer tacit knowledge even when it's not explicit in any docs, since it could abstract principles and concepts from the maze of data.

Chatgpt soon probably!