This last 12 months was a banger for AI because the technology went from area of interest to mainstream about as fast as anything ever has. 2024, nevertheless, will likely be the 12 months when the hype runs full-steam into reality as people reckon with the capabilities and limitations of AI at large. Listed below are a number of ways we expect that’s going to play out.
OpenAI becomes a product company
After the leadership shake-up in November, OpenAI goes to be a modified company — perhaps not outwardly, however the trickle-down effect of Sam Altman being more fully in charge will likely be felt at every level. And one in every of the ways we expect that to manifest is in “ship it” mindset.
We’ll see that with the GPT store, originally planned for launch in December but understandably delayed as a consequence of the C-suite fracas. The “app store for AI” will likely be pushed hard as platform to get your AI toys and tools from, and never mind Hugging Face or any open source models. They’ve a wonderful model to work from, Apple’s, and can follow it all of the method to the bank.
Expect more moves like that from 2024’s OpenAI because the caution and academic reserve that the previous board exerted gives method to an unseemly lust for markets and customers.
Other major corporations with AI efforts will even follow this trend (as an illustration, expect Gemini/Bard to horn in on a ton of Google products), but I believe it should be more pronounced on this case.
Agents, generated video and generated music graduate from quaint to experimental
Some area of interest applications of AI models will grow beyond “eh” status in 2024, including agent-based models and generative multimedia.
If AI goes to assist you to do greater than summarize or make lists of things, it’ll need access to things like your spreadsheets, ticket buying interfaces, transportation apps and so forth. 2023 saw a number of tentative attempts at this “agent” approach, but none really caught on. We don’t really expect any to essentially take off in 2024, either, but agent-based models will show their stuff somewhat more convincingly than they did last 12 months, and a number of clutch use cases will show up for famously tedious processes like submitting insurance claims.
Video and audio will even find niches where their shortcomings aren’t quite so visible. Within the hands of expert creators, an absence of photorealism isn’t an issue, and we’ll see AI video utilized in fun and interesting ways. Likewise, generative music models will likely make it into a number of major productions like games, again where skilled musicians can leverage the tools to create an unending soundtrack.
The boundaries of monolithic LLMs change into clearer
Thus far there was great optimism in regards to the capabilities of enormous language models, which have indeed proved more capable than anyone expected, and have grown correspondingly more in order more compute is added. But 2024 will likely be the 12 months something gives. Where exactly it’s unimaginable to predict, as research is lively on the frontiers of this field.
The seemingly magical “emergent” capabilities of LLMs will likely be higher studied and understood in 2024, and things like their inability to multiply large numbers will make more sense.
In parallel, we’ll begin to see diminishing returns on parameter counts, to the purpose where training a 500-billion-parameter model may technically produce higher results, however the compute required to achieve this could provably be deployed more effectively. A single monolithic model is unwieldy and expensive, while a combination of experts — a group of smaller, more specific models and certain multimodal ones — may prove almost as effective while being much easier to update piecemeal.
Marketing meets reality
The straightforward fact is that the hype built up in 2023 goes to be very hard for corporations to follow through on. Marketing claims made for machine learning systems that corporations adopted as a way to not fall behind will receive their quarterly and yearly reviews… and it’s very likely they will likely be found wanting.
Expect a substantial customer withdrawal from AI tools as the advantages fail to justify the prices and risks. On the far end of this spectrum, we’re prone to see lawsuits and regulatory motion with AI service providers that didn’t back up their claims.
While capabilities will proceed to grow and advance, 2023’s products is not going to all survive by an extended shot, and there will likely be a round of consolidation because the wobblier riders of the wave fall and are consumed.
Apple jumps in
Apple has a longtime pattern of waiting, watching and learning from other corporations’ failures, then blowing in with a refined and polished take that puts others to shame. The timing is true for Apple to do that in AI, not simply because if it waits too long its competition may eat up the market, but since the tech is ripe for his or her sort of improvement.
I’d expect an AI that focuses on practical applications of users’ own data, using Apple’s increasingly central position of their lives to integrate the numerous signals and ecosystems the corporate is aware of. There’ll likely even be a clever and stylish method to handle problematic or dangerous prompts, and although it should almost definitely have multimodal understanding (primarily to handle user images), I imagine they’ll totally skip media generation. Expect some narrowly tailored but impressive agent capabilities as well: “Siri, get a table for 4 at a sushi place downtown around 7 and book a automotive to take us” type of thing.
What’s hard to say is whether or not they’ll bill it as an improved Siri or as a complete latest service, Apple AI, with a reputation you’ll be able to select yourself. They might feel the old brand is freighted with years of being comparatively incapable, but tens of millions already say “hey Siri” every 10 seconds so it’s more likely they’ll opt to maintain that momentum.
Legal cases construct and break
We saw a good variety of lawsuits filed in 2023, but few saw any real movement, let alone success. Most suits over copyright and other missteps within the AI industry are still pending. 2024 will see plenty of them fall by the wayside, as corporations stonewall critical information like training data and methods, making allegations like using hundreds of copyrighted books difficult to prove in court.
This was only the start, nevertheless, and lots of of those lawsuits were filed essentially on principle. Though they could not succeed, they could crack the method open far enough during testimony and discovery that corporations would slightly settle than have certain information come to light. 2024 will bring latest lawsuits as well, ones pertaining to misuse and abuse of AI, equivalent to wrongful termination, bias in hiring and lending, and other areas where AI is being put to work without plenty of thought.
But while a number of egregious examples of misuse will likely be punished, an absence of relevant laws specific to it implies that it should necessarily only haphazardly be delivered to court. On that note…
Early adopters take latest rules by the horns
Big moves just like the EU’s AI Act could change how the industry works, but they have a tendency to be slow to take effect. That’s by design, so corporations don’t need to adjust to latest rules overnight, nevertheless it also implies that we won’t see the effect of those big laws for a superb while except amongst those willing to make changes preemptively and voluntarily. There will likely be plenty of “we’re starting the strategy of…” talk. (Also expect a number of quiet lawsuits difficult various parts of laws.)
To that end we will expect a newly flourishing AI compliance industry because the billions going into the technology prompt matching investments (at a smaller scale, but still considerable) in ensuring the tools and processes meet international and native standards.
Unfortunately for anyone hoping for substantive federal regulation within the U.S., 2024 is the 12 months to expect movement on that front. Though it should be a 12 months for AI and everybody will likely be asking for brand spanking new laws, the U.S. government and electorate will likely be too busy with the trash fire that will likely be the 2024 election.
The 2024 election is a trash fire and AI makes it worse
How the 2024 presidential election will play out is, really, anyone’s guess without delay. Too many things are up within the air to make any real predictions except that, as before, the influence mongers will use every tool within the box to maneuver the needle, including AI in whatever form is convenient.
As an example, expect bot accounts and pretend blogs to spout generated nonsense 24/7. A number of people working full time with a text and image generator can cover plenty of ground, generating a whole bunch of social media and blog posts with totally fabricated images and news. “Flooding the zone” has at all times been an efficient tactic and now AI acts as a labor multiplier, allowing more voluminous yet also targeted campaigns. Expect each false positives and false negatives in a concerted effort to confuse the narrative and make people distrust every little thing they see and skim. That’s a win state for those politicians who thrive in chaos.
Organizations will tout “AI-powered” analyses to back up purges of voter rolls, challenges to vote counts and other efforts to suppress or interfere with existing processes.
Generated video and audio will join the fray, and though neither are perfect, they’re adequate to be believable given a little bit of fuzzing: The clip doesn’t need to be perfect, because it should be presented as a grainy zoomed-in cellphone capture in a dark room, or a hot mic at a personal event, or what have you ever. Then it becomes a matter of “who’re you going to consider, me or him?” And that’s all some people need.
Likely there will likely be some half-hearted efforts to dam generated content from getting used in this fashion, but these posts can’t be taken down fast enough by the likes of Meta and Google, and the concept that X can (or will) effectively monitor and take down such content is implausible. It’s gonna be a nasty time!