For many, many tasks, you only need good enough results that accomplish a clearly-defined set of criteria. Smaller models that run 50x faster, in this case, could be far superior to a much slower model that meets the same criteria.
Sometimes you need speed; sometimes you need quality. There are very different use cases for each. For my own workflows, I sometimes want something very simple done ASAP; other workflows need "subjective" reasoning and careful crafting of responses.
>This makes these projects so much more trustworthy and easier to approach:
>> Were any of the words here written using AI? Nope. They came from my mouth or my fingers.
I have to push back on this a bit, as I believe (quite strongly) that we're shaped by the tools we use; text-to-speech LLMs are still LLMs, and generally their mistakes are shaped by the expectations inherent in their training. This, in turn, shapes the words that appear on the screen. For those who regularly use them, you then learn which word sequences are likely to be accurately transcribed, and this definitively becomes part of your thinking process. Over time, the LLM becomes tangled into your thinking; the use of AI, even in this way, very much can and often does shape the resulting words.
Anyone else ever think of somehow building an at-home carbon-removal system, for lower carbon content in the air of your living spaces? I know it's a silly idea, but one that has occurred to me from time to time.
A job doesn't usually involve a lifetime contract. And if it does, the severance required had better be incredible.
Nobody should "sacrifice future career prospects" just for a job. And if they do, it's hard to blame the employer on this, especially considering the premise implies they had choice in the matter.
Interesting! Would it be fair to say your company spend $100k to $150k per month on this?
Multiply this times many, many companies, and you can see how providing AI could theoretically be a good business to be in. Margins may be tight, though.
Also -- I'm convinced someone will figure out more use cases beyond software programming, which will result in many more companies spending $1k+ per employee per month.
It remains to be seen how much of this is a bubble.
If you say "Our goal is to represent the views of the right wing", then you're making it clear that you are being honest about your perspective. This then encourages both you and your readers to be aware of the bias, which does combat hidden influence of bias, making it more likely to be a choice.
This often happens when the news source is not honest about their biases, or if they don't make their biases clear. Fox News is the classic example: Their tagline is "fair and balanced," but their clear objective is to present the Republican/Conservative side of the story. This dishonesty about their bias is at the core of their credibility problems, though, to be clear, it does work in terms of effectively deceiving many people.
I think Fox News is a good example, because their public messaging has always been "fair and balanced" while at the same time blatantly have a bias; this is just one aspect of how they are clearly deceptive. If instead of calling themselves "fair and balanced" they said they were all about "the Republican Perspective on News" they would immediately be more honest, and it would be easier to understand them as an organization, especially for the people who are regularly deceived by them right now.
I'm not arguing that we should try to exaggerate our biases, or even to center them, but rather, we can become more honest by making our biases clear and explicit to those we're communicating with. Many organizations avoid openly addressing their biases, which makes them less honest overall, and more prone to being deceptive. If you're aware of your biases you can actually account for them, as opposed to letting them blind you. Further, if you're public with that awareness, others can account for them as well, and be less likely to be deceived (even accidentally) by your communication.
Too often, bias is ignored. It always exists. If we name it and make it visible, then we can have a chance at reducing its potential for deception.
I'm not saying that media should be used to influence instead of inform.
Rather, I'm saying you should acknowledge that you are influenced and will influence, and be explicit about what those influences are. This is the only way to actually combat bias; not by eliminating bias, but by making it visible, so it can be accounted for with everyone's thinking.
You can't be impartial; everyone has their own sets of biases that they can't get around. These are sometimes obvious, sometimes not, but they're always there. It's not necessarily intentionally dishonest to say you're impartial, but it fundamentally is dishonest to claim you're capable of the impossible.
Many people think you should avoid having bias. That may be the correct thing in some circumstances, but I think it's better to intentionally have bias, to make that bias explicit, and then to intentionally work within the framework provided by that bias. It should be open, public, and visible.
This allows for full transparency with the audience, increasing trust, while also giving a public "anchor" to guage your work against.
Many organizations do just this. Outside of news it's often just called "culture" or "branding," but it's more important, IMO, to be explicit, public, and clear about this in a news setting, and very much can serve as away to defend against outside influence.
Also keep in mind that none of the big chip purchasers (Apple, Nvidia, etc), want to depend on one supplier for their chips. There are huge incentives among them to encourage Intel's success.
Nobody benefits if just one company controls the state of the art in chip manufacturing, and Intel is one of maybe two other companies positioned to have a chance at competing effectively with TSMC.
Cloudflare is increasingly a problem in terms of blocking huge geographic regions, often without the website operators even being aware this is happening. All in the name of "security."
I actually really like Composer 2. For my use case, between the planning tool, and getting it to ask a lot of clarifying questions, I regularly get very good results. I'm not doing anything complex though; mostly staying in the lane of very common web app type code.
I really want to like the Kobo. I really do. But I've had such bad luck with their devices. For example, sometimes the pages randomly start turning, really fast, so I completely lose my place. It also never reliably syncs between devices. And the integration with Overdrive is unreliable, only working some of the time. I also read it in the bath sometimes, which supposedly is one of the features available due to the water resistance, but the steam causes random clicks on the device, which makes it not really functional.
For me, I've mostly switched to reading on my phone. Dark mode, plus OLED, works very well for my needs.
I too have noticed the shift in completely different contexts. Definitely gives me real pause. Mental acuity and sharpness is so important; it's the foundation of who we are as people...
Has anyone else noticed this, as they've scaled up their AI coding use? I've found it harder to stay on task, and it's affected a broad range of my personal activities. I'm able to make incredible things happen with AI tools, but do worry about the personal costs.
Sometimes you need speed; sometimes you need quality. There are very different use cases for each. For my own workflows, I sometimes want something very simple done ASAP; other workflows need "subjective" reasoning and careful crafting of responses.