I'm one of those senior devs with 20+ years. I've been a bit lost but I agree with your predictions about our demographic. Thanks for a great vid. 👍
@TreantBG
9 ай бұрын
Solana already flip BNB briefly. So this is expected. And btc at 50k is maybe conservative.
@dawid_dahl
8 ай бұрын
The moment you showed your successful vs failed predictions I subscribed. 👌🏻
@indydevdan
8 ай бұрын
Glad to have you onboard. Self accountability (although difficult at times) pays massive dividends.
@teddyfulk
8 ай бұрын
Love the idea of prompt based interview questions
@quickcinemarecap
9 ай бұрын
00:02 Predictions for 2024 in Advanced AI engineering 01:52 Position yourself by learning relevant technology for future success 05:41 In 2024, fundamental engineering skills will be crucial for knowledge workers, with the chat interface becoming ubiquitous and a shift towards new user interfaces for LLMs. 07:46 Mastery of TypeScript and Python ensure employability in tech industry 11:36 Technological dependence and AI impact predictions in 2024 13:35 OpenAI continues to lead the AI race, facing legal challenges but progressing with next-generation models. 17:24 Mid-levels will see specialized job roles for prompt engineering emerging. 19:13 Focus on becoming an agentic engineer 22:46 Python and TypeScript are becoming dominant for AI tool experimentation. 24:37 Building agentic software for AI engineering You Can't Judge A Video By Its Cover. you can by its first few chapters and certainly by its last.
@agchain6554
8 ай бұрын
Hello, I like your videos! Awesome, you give a lot of knowledge. I'm new in this space, just learning Python, and how to use it to create those AI agents. And I think this is the way to go; there will be a shift where people have to automate a lot of things to compete with others who already did that!! And for the prediction, I think BTC also will close less than its ATH, so around 50K-55K.
@realkiddshady
9 ай бұрын
Another great video. Thank you!
@IslandDave007
9 ай бұрын
One use case that is missing but I think is a great opportunity for 2024 is RAG - private "chat" with your corporate documents using local LLMs inside your network. The current LLMs are good enough for this right now and the vector DB tech and user interface options are looking good - examples include AnythingLLM and Cheshire Cat. Leverage Docker for hosting, Ollama or LM Studio, etc.
@pmiddlet72
8 ай бұрын
Thank you for all of your fantastic posts! I find your work some of the most valuable out there (in fact I'm quite interested in your current project and already fully on board with multi agent software that does actual work rather than draw melted psychedelic pictures of bored monkeys - I'd actually be interesting in chatting with you offline if that'd be of interest - I have some specific domain use cases I'd like to run by you). I think most of your predictions are on-target. But I'd like to dive in further if you don't mind - so some q's/comments. On prediction hits/misses: I dunno if 'anything' above 50% will make anyone win big. Given error margins, even hit/miss proportions of 45-55% could conceivably be roughly equivalent to random outcomes. It's hard to dichotomize predictions into right/wrong, yes/no, true/false etc. I look at them as governed by probability, so in my world things can be more of a 'spectrum' of rightness/wrongness, particularly when predictions are conditioned on beliefs (and their nuances) about them. Where do you think the golden threshold for hit percentage, if it exists, would be for 'winning big'? On AI sentience: I wasn't entirely clear on this. Are you simply stating that there will be a 'scare' on something that appears sentient? Or rather that emergent properties of sentience will be seen in generative AI this year? Although I don't agree with everything he writes/states, I'm with LeCun on this one - I don't see sentience, in the way we have at least some definitions established, coming out of GAI alone. There are more cognitive/psychological elements that are required for much of the known definitions of sentience to be satisfied. I do agree that there will prob be another 'scare' though (or at least the hype of one). And I can't wait for the clickbaiters to come out of the compost heap on this one .... Fundamental skills vs shortcuts: Absolutely agree here. However, do you think that businesses may be none the wiser on what is needed to support and advance the tech in a value-driven manner? I would predict that businesses will 'flock to the tech' long before a proper strategy for using and implementing it is thoughtfully and expertly formed. I've seen this far too often in larger corps with whom we consult (spoiler alert: they ended up with frighteningly large sunk costs). So to reduce risks and increase value with such tech, your call for the need of fundamental KSAs in addition to AI skills (prompts, experiments, etc etc) should be a clarion call to all institutions, large and small, who are about to take the AI leap. Moving up the stack for senior engineers and up: I think the point of my above question/comment can be the thing that stifles some upward movement. Many in big tech look at the '$400k+' days being over. I mean some institutions 'get it' (wrt to the value of the expertise) and respect their seniors/leads/principals. Others think a horde of cheap junior labor will be 'good enough' with the tech. These are those 'tech first' companies I'm referring to above. And I think they're quite abundant. Cancelling a recession: Do you believe we will honestly ever know who or what slashes a bona fide recession given we are now grappling with the definition of what a recession even is? My personal belief is, from a measurement PoV, we measure the economy quite wrongly. The SPX, GDP, jobs numbers, and inflation rate are certainly important elements to this, but as part of a hierarchy, even together they hardly paint a complete picture when one considers various economic classes of citizens. For example, despite the fed slowing interest rate changes in recent weeks, price gouging is an ENORMOUS problem. The number of middle class folks losing disposal income is increasing. And of course there's the '3rd wave' of M&As which look great for investors (it's good to be the king), not so much for the consumer. I look at all of these factors and just empirically, I don't think the solution to whatever the fed ends up defining as recession will be borne on the backs of the M7 (although I do believe that advancements will change the way our economy works in some ways - bearing more startup / smaller biz fruit with those who are truly engaged). SPX: This was an insightful callout. SPX (and other measures) do go up in election years sure (the usual artifact of elections). I'm curious though, given the rather diametrically opposed candidates out there in party land, whether poll result edging to one side or another before voting begins will begin to sway the 500 to a downside this year, even if it does percolate to a high. Thoughts? Big tech margins: 1000% on this one. Although due to cloud DB (Snowflake becoming quite infamous these days), and other 'Cadillac cloud' sticker shock that companies are trying to control, is it more likely that the market will swing confusedly to the tidal wave of smaller companies with '.ai' as their URL domain extensions? Of course there's the question of which of those small companies will be bought up by one of the giants.... this should be an interesting ride. Crypto: This one's always a mixed bag. But traditionally, retailers both in trad stocks and crypto, with the exception of the very quant savvy, knowledgeable, and lucky few, begin dumping crypto only for the institutional investors (I^2's)to sop up the blood stains. By the time an IPO/ICO even occurs, with few exceptions, it's generally understood that I^2's often already hold anywhere from 75-90+ % of shares by the time investors even get to the punch - and then ride the manipulation waves. I find the BTC ETFs interesting, and with those, yeah, the price could conceivably sky to $50k before correction (of the distribution of ~150k simulations I've been toying with, that number doesn't seem outlandish - but I've been trying to hunt down proxies for priors that most resemble the past 24 month - and particularly when ETFs are intro'd). What's interesting in 2023 is that, at a daily level, BTC and ETH have been trending positive quite linearly in step with the trad markets and indices. As far as 'decoupling' - I dunno yet. If the power of the USD for the consumer and consequently DPI push the gas pedal on the downslope, I could envision the trad market moving oppositely from crypto (naturally with volatility), but not quite a definitive decoupling. I'd like to listen to what's behind your thoughts on this.
@jeremybristol4374
9 ай бұрын
I feel like most of these are spot on. UX is due for disruption, as all companies and all products gain access to AI-assisted experience development and AI-augmented experiences. Explainable AI will become more important as a way for companies to adjust weights to improve outcomes. However, this democratization of experience quality may increase the need for human-in-the-loop; if so the people managing product experiences will have greater reliance on sociology and psychology and less on UI trends.
@kora5
5 ай бұрын
In your blog on 2024 preductions, please add the link to this video so people can choose to come watch here instead. (I linked to your blog, and wish people could watch instead of reading)
@sevilnatas
9 ай бұрын
There is no such thing as a national debt bubble as long as the US dollar stays as the predominant world currency. Unless BRICs is able to come up with a currency that can surpass the dollar, or the US collapses or at very least declines drastically, which many of your predictions seem to indicate against. The national debt concept is a myth perpetuated by lobby captured politicians, as a means to the end of the short sighted capital class (see MMT). I worry more of a decline of the nation states, in favor of a corporate run world. Based on the rise of AI probably being most important advancement of the future and if your predictions are anywhere close to correct, it looks like the corporations will control them and be in the best position to run the world.
@remsee1608
8 ай бұрын
This bro hasn’t heard of cryptocurrency yet
@him4440
8 ай бұрын
Well said
@sevilnatas
8 ай бұрын
@@remsee1608 This bro hasn't seen crypto amount to anything beyond a gold replacement yet. After several years of operating a medium sized bitcoin mining operation, it has proven to both be inconsequential on the international level and merely a highly speculative commodity-like wealth parking instrument on a individual investor level. That is for bitcoin and to lesser extent eth, the rest of the cryptos seem to just be rug pull foder, which tarnishes all of crypto's reputation with the general public.
@sevilnatas
9 ай бұрын
If countries like India and Brazil were smart, they would start to evolve their education system towards a higher mix of liberal arts into their predominantly technical post graduate system. They made the brilliant move to emphasize under graduate and graduate level technical training, in their universities, and limited the liberal arts requirement, in order to hyper accelerate the countries graduates into the job markets of the west. They lack a fundamental level of creativeness that western graduates have historically been given in the western education system, but has been declining, due to short sighted thinking. If prompt engineering is the future of software development, software development will be more akin to short story writing, than code development. A developer will be better off understanding big system architecture and the best way to communicate to the AI their intent, via the written and probably eventually the spoken word.
@Canna_Science_and_Technology
8 ай бұрын
Open source LLMs will be at par with GPT 4. 100%
@mitmaxim
8 ай бұрын
Do you mind sharing channels/ people you learn from? TY
@hroosky
8 ай бұрын
Assuming your predictions for local LLM performance prove accurate, can you make a prediction for the optimum developer laptop/workstation spec?
@JustJervon
8 ай бұрын
Bro, you have a discord or some way to follow and seek advice?
@yvrary
2 ай бұрын
damn the prediction for Anthropic Artifacts was damn right!!
Пікірлер: 26