Reading through the specs I was pleasantly surprised to see Wacom EMR on the list, I enjoy it on my laptop and seems like it would be a great choice here. Best of luck to you and the team!
I do think OpenAI has a point in what they're saying: if we expect human-level competency of AI, it needs to be able to see and train on human-accessible content and ideally with a similar distribution.
For example, I make an open source Firefox web extension for filtering internet content with my own classifier. That literally would not be able to exist without being able to be trained on web content, much of which is copyrighted. Requiring that I somehow either a) use only attributed data or b) detect and not use copyrighted content when trying to build something representative of my source distribution (e.g. the web) sounds like a recipe for a poor outcome. Now maybe my addon isn't your cup of tea - but what if you found out that the next generation of uBlock Origin etc. could not be as effective because of legislation because it wanted to use an AI model? Legislating too heavily around this area will, I believe, have a tremendous chilling effect for small businesses and open source folks trying to innovate in AI.
I've also worked commercially in the creation of two closed source machine learning models, but the domains were restricted enough that web content was not a particularly helpful input. One did all right, and one did not. Seeing bets succeed and fail gives me appreciation for the long-term and uncertain bets that OpenAI has been making for ages finally coming to fruition. I think without businesses being willing to make those bets the GPU-hours would have been hard to pay for.
I've wondered if potentially a different way out of this is not restricting the use of copyrighted material in the training process itself, but rather to instead only consider the created final works. Of course there are thorny problems there, too, but I don't see that having the same chilling effect on research and probably a lesser effect on business as well. One thing I think is clear though: we've reached a tipping point in the US similar to 1998 when the DMCA was legislated where the technology is forcing us to think carefully about what copyright means.
So I have question for those on HN who have meaningfully worked in the creation of not just AI-generated content, but in the creation of some AI model that others use freely or commercially: what seem like promising paths forward here?
Or to those working in copyright law (like @williamcotton): how do you see the status quo and potential paths forward?
> I do think OpenAI has a point in what they're saying: if we expect human-level competency of AI, it needs to be able to see and train on human-accessible content and ideally with a similar distribution.
Any human who wants to access copyrighted works is by law required to honour the copyright - whether that means purchasing of licenses, timed rental access or immediate cease and desist of usage. Why should AI (and the billion-dollar-backed companies building them) get different treatment?
Or to turn it the other way: if the billion-dollar-backed companies building AI models do get a free pass then surely humans should too?
I think there's an important distinction here though. We access copyrighted material all the time though, as accessing copyrighted material is not always protected in the ways you describe. We view copyrighted images via Google Images, for example. That works because Google stores metadata to point at the content and then loads it. Copyright is (broadly) more about the "not copying" it part.
>We access copyrighted material all the time though, as accessing copyrighted material is not always protected in the ways you describe.
sure, because there's an incentive to attract humans to view such content. It's advertising. Google images isn't built out of goodwill, but is now a target to optimize for to maximize human traffic to get human eyeballs to view ads (or paywalls) for humans consume more products. Having a bot come in ruins that, and the literal billions thrown at adtech to try to verify organic traffic shows that these bots are not desirable metrics for those who pay for ad space.
That perspective from a business lens shows the difference between a human viewing copyright material and a bot. Humans are monetization targets, bots are not. Humans can advertise for you to other humans, bots can't (well... not yet. But do we really want to be talking about ChatGPT ads this early in the LLM era?)
I'm really excited about this! I know the Keras team has been hard at work and it's cool to see the 3.0 release as sort of a reborn framework. One thing I'm particularly stoked about is the ability to load models from other frameworks but then work with TF and Keras based exports for productionization, especially on e.g. edge. I think there's a lot of goodness beyond just being able to pick and run a pre-existing backend. Thanks for all the hard work!
I thought this line was gold: "And I worry that being immersed in a world of hungry, gaping maws, grasping for ideas – any idea – to gobble up; to compulsively transact in ideas for status, money, and friendship; pulls me further away from the divine somehow. Chasing ideas can be just as materialistic as chasing money or power."
Hello!
I can see you're discouraged. Have you had a chance to sit down in person and talk with a fellow programmer friend or colleague about your experiences and what you can do to change your situation?
I am less pessimistic about the job outlook. As both a developer and a hiring manager, I do not think the need for junior developers goes away with AI. Managers aren't magically all just writing code by whispering into the AI and having it produce their product.
A few questions/thoughts; please keep in mind I do not know the ins and outs of your situation:
- You've indicated you're sending out endless applications. As a hiring manager, one thing that particularly stands out is when the applicant shows me that they are interested in my company's specific job posting - so quality over quantity may be key. Are you including a cover letter that shows you've looked up the company a bit and are able to make a convincing argument you'd be a good fit for the job? It might seem old-school but the number of applications I see with a well-sharpened one page resume and convincing cover letter are less than 5%.
- Have you sat down and reviewed your resume with a hiring manager to help sharpen it up?
- If you are getting as far as rounds of interviews but are consistently getting rejected, it may be worth it to double check your references. Perhaps one of them feels differently about your performance than you'd expect, and is burning all of your interview hopefuls. (I encountered this on the job just this past week.)
- Have you sought and received helpful feedback from job applications where you did not succeed? Ideally this sort of feedback can help keep you from painfully repeating things.
- Are you willing to work onsite rather than remotely? Are you willing to relocate someplace less glamorous? If so, I think this may open a few new doors, and may provide opportunities to build experience.
- Regarding starting your own business. You've mentioned that you do not have a burning passion to be programming all the time. If you are considering starting your own e.g. consultancy, I'd like to note that - successful or not - one thing I've seen as a strong common factor in all entrepreneurs is that they have a strong passion or dedication to their mission and often their craft.
- Switching career paths. I do not know your situation, skills, or true expertise - but your post sounds as though you are deeply discouraged. I'd recommend seeking counsel from family or friends who know you well - they may not know software development well, but they may be able to provide perspective.
Can you tell us a bit further about your experiences and specifics on things that have worked or not?
At my workplace, I started in software and then was able to transition into embedded software about 6 years in. I credit a lot of that to a) working at a place that had both the field I was already in and the field I was getting into and b) the company culture/supportive managers/current development needs. Some fields I think this could work, others not so much...
You might get some more targeted responses here if you could give us an example of the types of roles you're applying for. As noted some fields are tougher too so I think general advice is only going to get you so far and specifics may help in so far as you are able to share.
But as a member of a small-ish company, I remember there was this time the sales VP insisted that we MUST have this product out by a certain tradeshow. Engineering worked hard and delivered, and although the thing had warts, it gave our company an advantage: the current market leader showed up at the next tradeshow a couple months later (e.g. "20%") with a similar product and price point, but our lead in time meant that we were able to exist as a strong competitor. This single product launch and subsequent sequels lifted our company for quite a long time.
So now when the sales guys insist we might need something by a certain date, I carefully evaluate rather than dismiss out of hand - I've learned to trust their role as well.
I'm a big proponent of state machines. They are an excellent tool for unambiguously communicating how a certain critical part of a system should work - to that end, they are not only a coding tool but also a social one to get everyone on the same page. I've wandered several times into a situation where the state is not being well-handled and while it's not often loved, getting the team to hammer out a state machine does wonders for system reliability.
It's true that they have limitations and are not always the right tool, but it's often valuable to realize that you implicitly have a state machine already whether you wish you did or not.
Same here. I am a big proponent of state machines. I have found that state machines bring sanity to an otherwise complex situation. For example, most inexperienced developers will check presence of certain fields to imply where someone is in the process whereas it would be better written using a state machine.
Something that could read the screen at any point in time, across apps and flag/filter NSFW images, all running with a local model and no outgoing internet connections would be pretty neat.
Unfortunately my experiences with mobile app stores put that likely in the "not possible at all" category rather than the "tricky" category like it might be for desktops.
