The real challenge of AI is democratisation.

We are in middle of a boom in AI. While it may seem like a sudden explosion in the media; this moment has been building for 15+ years now. From the birth of the internet, to the huge explosion in online information, to algorithmic feeds, then voice assistants, and now at Generative AI/Large language feeds - every invention has grown on the bedrock of technologies before it.

Of course, the fervour we see now is from people experiencing a step change in technology. Till December 2022 you had to write eassy using your own brain and hands; come January 2023 you just provide a small prompt and voila! ChatGPT will write a whole essay for you. WOW.

But we’ve played this game many times before:

• New technology is available
• The technology gets hyped and people are promised a better future
• Technology gets widely deployed
• In the process of societal adoption, it gets co-opted by
• Eventually, it gets completely owned by . The new boss is the same as the old boss. The average person is nowhere better.

One interesting aspect of computers is that they are comparatively much more accessible; order(s) of magnitude more accessible.

The steam engine was firmly on the capital side of the equation; it powered everything from trains to factories but it was a while before mechanisation entered the home and became accessible to the average person. Yet overtime we’ve given up on things like sewing machines and become strictly consumers.

Computing and the internet sit firmly in the middle. Almost by definition - every laptop & smartphone is the “means of production”. The entire “internet” however is but a handful of companies; yet because of its inherently decentralised nature; everything from Mastodon to torrents and more can and do exist.

If it wasn’t for that design; the internet would like the TV - with much less control in the hands of citizens.

The challenge with AI is to keep it’s capabilities as democratically accessibly and user programable as possible; with the internet being the minimum bar. Thankfully with models being open sourced and even run on Raspberry Pi’s, we are off to a good start! But we must remain watchful and make sure this happens.

People of my ilk (technologists) are gung ho about applying AI to anything and everything - but this question of democratisation is much fundamental and its effects more pervasive.

Hey hey hey, who randomly went to UCSC to talk to students about tech careers. Public speaking is easy for me, the drive back was the crazy part :D

“Jokes of the proper kind, properly told, can do more to enlighten questions of politics, philosophy, and literature than any number of dull arguments" —Isaac Asimov

Thank you to everyone who got me to 100 likes!

Wow 😂 - this is an auto-generated post from Tumblr. Thanks to the 37 bots and 2 humans who got me here.

The Axe

It was a melancholic start to the year. It rained incessantly outside and the mood was grey. Then came the axe in the form of layoffs at “BigTech” firms with Google doing it’s first major layoff ever. I was emotionally at-sea for 2 weeks as I reeled from the new of the Microsoft layoffs. As a 31 year old with sufficient financial savings and no kids or debts/loans hovering over me, I would be fine. But I’m also an immigrant who existence depends on having a job here and who had spent a lot of effort/time moving things around to get here. It would be annoying to put it mildly. I’m just happy that I coped by watching TV and reading books; and didn’t take up my usual route of getting drunk/stoned/spaced out.

It is a tough time and I directly know multiple people who have been affected. One silver lining was the report that most tech workers are landing a job within 3 months. So this is more of a displacement that a destruction of tech jobs. They’re moving away from more speculative areas to places of clearer ROI/need.

Layoffs also brought attention to economic iniquities with the tech industry itself with CEOs getting multi-million-dollar pay packets and no direct consequences while other bear the consequences of their decisions. People frame this as merely a moral problem, but it’s a public policy one. France and Germany make it hard to layoff workers which means the wanton hiring that’s triggering this is less likely to happen in the first. Of course, there is always the American claim that this makes industries “less dynamic”. Ah, hiring and firing are but key sources of dynamism in an organisation not the quality of its talent, it’s psychological safety or the ability of its workers to focus on their work.

It definitely felt horrible to realise that the big tech layoffs were largely financially motivated and felt opportunistic. Even accounting for economic contractions, the companies still earn billions in profit and pay dividends to shareholders. They face no chances of bankruptcy and could take a hit in profitability for a few years till they reallocate their human capital. But that would hurt the shareholders who are also common citizen and have financial interests in the stocks staying high. I don’t mind this financial book rebalancing per se as long as people have safety nets; we should aim to minimize the suffering and disruption involved.

ChatGPT

ChatGPT took over the airwaves and I (like many others) was genuinely impressed by its capabilities.

“Any sufficiently advanced technology is indistinguishable from magic.“ say Arthur C. Clarke; and that’s truly what we have. After many years of research papers and demos, the world has access to some really cool new technology. The last time I was wowed similarly was when Google demoed Duplex. The excitement all around is palpable.

Creatives, technologists, and designers are excited by the possibilities.

Engineers are happy that they have something other than a new Javascript framework to look forward to.

Founders and Venture Capitalists are excited they have a new buzzword to add to their pitches and fool their LPs with.

All in all a good moment.

Personally, this feels like a truly transformative technology to me that will change a lot of worflows.

It’s easy to see tech like this

• Writing my emails, documents, and more.
• Helping me compose things by providing inputs and bouncing off ideas
• Summarising information
• Enabling more programmatic creations for infinite games and custom movies

It is not as behind-the-scenes, as say nuclear fusion. It is also much more real and obvious than "blockchains” - can’t believe we had to suffer through that.

Randoms

I watch a lot of Veep and am almost done with the show. It’s a British show in American shows clothing. I also got really confused by my insurance claim payment process for one hot moment.

Me and Namrata spent a lot of time looking for a new place to rent; zeroed in one one - only to realise that our lease ends a month later than we thought. All this effort wasted uuuggghhh.