avinoth

Probably too much of the training data that has Magic wands were from a video game.

Because humans have basic needs like food, shelter and they can be put it jail. And they have loyalty, patriotism, feel part of the society. Those threats & incentives have made it for a good society to function so far.

I think that line is getting blurrier by the day with the current administration. The companies very quickly fell in line before even any consequence of note. Donations to the ballroom, calling it super intelligence, renaming the gulf, etc.

Would you really trust that OpenAI wouldn't skew the results to tend to a certain way or that Grok will always be neutral in how it operates?

From a third world, there was a time when US products & companies were more trustworthy, I agree. But, there has been enough display from these Cos to show that they have lost that badge.

But in Twitter or Facebook, the complexity is not in the app but with the scale. How do you support the network, graph, querying, fetching, etc. That is not the case for a self-contained binary like Claude code CLI that talks to the same API irrespective of 100 people are using or 1M people are using around the world.

I don't think it does either, but it's surely overvalued and have to sustain it's ridiculously high forward P/E. Hence the semi-regular news about FSD, TeslaCab, Optimus Bots, and now Flying cars. Seems it's working. It's always priced in with the "potential future".

Yeah, true. It’s sad to see the Market forces being nothing more than a whimper on these massive untenable mergers.

Exactly. What they’re doing, repeatedly, is setting a narrative that this is how software engineering should be. Always be reliant on tokens. Don’t worry about uptime or errors - agents can address it, don’t worry about security - agents can address it, don’t worry about code comprehension - agents will always be there for you.

This is like one of Google’s infrastructure blog. Google scale things are only relevant and appropriate for them. We had lots of bitter lessons from smaller engineering teams following the process with the justification that “this is how Google does it”.

Similarly, Anthropic can afford to do this because they have virtually unlimited tokens and labs to train models for their specific usecase. The rest of us need code comprehension lest we become dependent on continuous supply of tokens (which is a state these labs badly want to achieve)

So, unless you work at a lab, or have unlimited tokens guaranteed through the life of your company, it would do both the company and you good to continue to read the code and not let these 30k PRs per engineer be the norm.

An excerpt from the article makes my point

> Nobody on the team keeps a complete mental model of Claude Code in their head, and we don't think anyone needs to. When someone wants to understand a part of the codebase, they ask Claude to write an explainer for exactly that slice. It takes minutes and reflects the code as it is today.

Mark my words, these thrusters and flying car stunts will eventually be used as a justification to merge Tesla into SpaceX, the same way how they justified space datacenter to merge X.ai onto SpaceX.

There was optimus robots already on Tesla side and inference / r&d funded by x.ai (now spacex).

One thing I've observed Gemini being the best is in the context of non-western stuff. Like, when I ask about obscure srilankan food, or chat about Indian/Tamil movies or songs, Religious customs & beliefs, Gemini almost always answers accurately and very close to how helpful they are in Anglosphere related topics. ChatGPT also does do a bit better recently, but Claude/Grok are absolute trash on those fronts.

Google's years of research project into Indic scripts, image trainings, translation efforts & transliteration efforts is literally paying off. The AI adoption from non-English, non-Chinese population will only if you have these contextual localisations, which I think Gemini is poised to do well.

For example, you can ask for a shopping list for a marriage function in a town from Tamil nadu (the marriage custom varies by districts), and Gemini provides contextual and helpful answer (like what is the custom followed there, what is gifted traditionally, etc), whereas if you ask Claude/ChatGPT, it talks about marriage registeries or pretty generalised answers which is not relevant or useful.

Is it only me or is the changing background disturbing for others as well? Maybe it's my ADD brain finding the hook, but it takes the focus away from the content whenever it shifted.

The quickest sign I have right now to find if the developer/team don't read their code is looking at their recently AI-generated test cases. Often, coding agents go too verbose in their tests and they are most often overlooked even during normal reviews. If there are verbose tests, like declaring a constant and asserting it's value, you know that the code isn't being read much.

I moved the opposite direction. Few months back, I was absolutely sure that the days of me being a software engineer are numbered. The programming I used to love and fun was gone. But, as I got to work more with Agents, factory & the infrastructure surrounding it, I'm more optimistic about the kinds of problems we could tackle with this, that wasn't possible/cost-effective before. We've only barely scratched the surface with what AI could be leveraged in the real world.

I'd say if it's both doing it, it's a strong signal for the opposite.

Is it possible to do that through Veracity? I don't want to use a non-Veracity based solution when Veracity already supports it.

Comic aside, I can’t help but wonder how come these one-off websites are all using .ai domain when at the latest check they cost $80 per year with minimum 2 year commitment. So $160 for a comic that is at best few news cycles away from redundant, and at best a recurring comic series about AI..? Why not just use a cheap tld like .fyi or .art, or something.

Here I am trying to find appropriate AI prefix or suffix so that I can continue using .com , while these pet projects are all flaunting .ai domains

I said it before, and I'll say it again. The winners of the AI race are going to be the ones who control the products to deliver the AI features to the end users. The obvious ones are Google with its myriad of consumer tech, then we have Apple, Meta, Microsoft, etc. That's why OpenAI, Anthropic are desperate to build the applications themselves than being just an API provider. Meta is better off focusing on such initiatives than trying to complete with the labs on the model strength. They seem to have realised that finally, so that's refreshing. Whether the solutions they bring are useful or not is a different question, but this is definitely how you reach the end users, not with the strongest models.

This is in line with what I was discussing with my friends recently. Governments across the world weaponised their border protection for all their invasive data collection citing terrorism and other threats. This more or less started post 9/11, and used that as the reason. Now, the AI agents and their hacking is exactly going to be used as an excuse to introduce the real ID verification they've been trying to implement for a while now

I recently wrote a short blog post about exactly this worry. Link if anyone's interested https://www.vinoth.net/internet-for-humans/

It is not an allegation that they've faked it, or that there is no risk of rogue agents. I'm alluding to the scenario where the supposedly illegal/immoral activities are rewarded with increased valuation and funding as long as they are done with AI Agents rather than curtailed & punished.

Right after the HuggingFace incident, we all saw how every company clamored to claim how their AI models have also broken out of Sandbox, and hacked some stuff. As ridiculous as it was, they all used that to demonstrate their model's capabilities.

Rookie move. Missed the chance to claim that an AI Agent swarm hacked it autonomously and claim billions of VC investment.

> RubyGems disables new user registration

> On May 16th, registration with disposable emails was disabled as well.

These kind of repeated attacks or attempts to attack by agent swarms is only going to make the experience worse for the rest of us actual humans. ReCaptcha is already annoying enough, I can’t fathom what comes next.

Unfortunately this makes a perfect justification for governments and companies to push for real ID verification.

Yeah, and its extra funny with everyone agreeing Dave should own it, but he isn't on the email loop at all.

Yeah but the contention is here automating the bookings. Did Claude book the flights, hotels, tickets and attractions for you or would you let it without oversight?

Which is good if sailing is what you wanted and like to do, which is what the author and the OP is implying. You don't necessarily have to "succeed" or "reach" somewhere.

I’ve had similar experiences. I now have an explicit line in AGENTS.md to not use subagents unless explicitly requested. It also helps that for the tasks that are big enough to benefit from subagents are also the ones with high chances of going off-rails and/or a poor review phase. I’d rather do the orchestrator role and that way I can split up the review phase in a much more manageable chunk.

I'm beginning to build/host a series of own-use, self-hostable products for personal benefit, starting with an LLM chat for me & my wife, a photo manager, our personal calendar, "whitelisted" youtube app for kids in our home, investment tracking, story/music playlist for our kids, and so on. Plan is to run and maintain our own ecosystem of essential products for our family.

> Ironically, the biggest multiples it’s generated for the industry have been from a low-dazzle text forum written in Arc Lisp.

Wholeheartedly concur. I'm yet another datapoint for this statement. I picked up programming, picked up a tech stack, found jobs, and launched profitable sideprojects, all with the help of this site.

No, please no. Hyperlinks exist, and they work just fine. HN is one of the last few sites that doesn't glamorize the UX (nothing wrong with that). I'd rather read a thousand words than a picture.

I guess it's a bit of this and bit of that. The products I've sold before or discussed with others were on the former side. I think that was largely due to the metrics that are usually used to qualify (churn, MoM growth, etc.) are bit less impactful compared to a mature, stablized product.

It's mostly seen as a jumping board for them to quickly get going, compared to finding idea, building it, getting some SEO footing, etc. That's just my anecdotal experience btw.

From my experience, x arr is not a reliable metric for products of this size. They don't exactly have risk covered returns due to their size and they're usually bought for their growth prospect than any meaningful return it can give.

I didn't have any particular expectation per se, any x arr with the potential baked in is fine. But the most conversations I've had are bog-standard x months of profit and people looking to acquire it as an asset, which makes sense for them, but doesn't for me as I mentioned before the risk is pretty low in keeping it myself.